diff --git a/content/post/2018-iran-protests-pt2.md b/content/post/2018-iran-protests-pt2.md index 07821b251..c5d54d489 100644 --- a/content/post/2018-iran-protests-pt2.md +++ b/content/post/2018-iran-protests-pt2.md @@ -2,7 +2,7 @@ title: "Iran Protests: DPI blocking of Instagram (Part 2)" author: "Leonid Evdokimov" date: "2018-02-14" -tags: ["technology", "iran", "DNS", "DNS Hijacking", "TCP injections", "DPI", "country-ir"] +tags: ["technology", "iran", "DNS", "DNS Hijacking", "TCP injections", "DPI", "country-ir", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/2018-iran-protests/index.md b/content/post/2018-iran-protests/index.md index 186007403..29b41748d 100644 --- a/content/post/2018-iran-protests/index.md +++ b/content/post/2018-iran-protests/index.md @@ -2,7 +2,7 @@ title: "Iran Protests: OONI data confirms censorship events (Part 1)" author: "Maria Xynou, Arturo Filastò" date: "2018-01-05" -tags: ["iran", "censorship", "country-ir"] +tags: ["iran", "censorship", "country-ir", "theme-im", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2019-benin-social-media-blocking.md b/content/post/2019-benin-social-media-blocking.md index b6423dab5..4560597c4 100644 --- a/content/post/2019-benin-social-media-blocking.md +++ b/content/post/2019-benin-social-media-blocking.md @@ -2,7 +2,7 @@ title: "Benin: Social media blocking and Internet blackout amid 2019 elections" author: "Roderick Fanou (CAIDA, UC San Diego), Ramakrishna Padmanabhan (CAIDA, UC San Diego), Arturo Filastò (OONI), Maria Xynou (OONI)" date: "2019-04-30" -tags: ["benin", "censorship", "country-bj"] +tags: ["benin", "censorship", "country-bj", "theme-im", "theme-social_media"] categories: ["report"] --- @@ -38,41 +38,41 @@ In the following sections of this report, we share [OONI](https://api.ooni.io/fi ### OONI measurements -OONI [measurements](https://api.ooni.io/files/by_country/BJ), testing the accessibility of websites and apps, have been collected from multiple networks in Benin since 2017. OONI’s [Web Connectivity test](https://ooni.io/nettest/web-connectivity/) is designed to measure the TCP/IP, HTTP, and DNS blocking of websites, while OONI’s [WhatsApp](https://ooni.io/nettest/whatsapp/), [Facebook Messenger](https://ooni.io/nettest/facebook-messenger/), and [Telegram](https://ooni.io/nettest/telegram/) tests are designed to measure the reachability of those apps from local vantage points. +OONI [measurements](https://api.ooni.io/files/by_country/BJ), testing the accessibility of websites and apps, have been collected from multiple networks in Benin since 2017. OONI’s [Web Connectivity test](https://ooni.io/nettest/web-connectivity/) is designed to measure the TCP/IP, HTTP, and DNS blocking of websites, while OONI’s [WhatsApp](https://ooni.io/nettest/whatsapp/), [Facebook Messenger](https://ooni.io/nettest/facebook-messenger/), and [Telegram](https://ooni.io/nettest/telegram/) tests are designed to measure the reachability of those apps from local vantage points. The following chart, based on [OONI data collected from Benin](https://api.ooni.io/files/by_country/BJ), illustrates the blocking of social media sites on 28th April 2019, amid Benin’s 2019 parliamentary elections. ![](/post/2019-benin-social-media-blocking/OONI - Blocking of social media in Benin.png) -**Figure 1:**Blocking of social media in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ)  +**Figure 1:**Blocking of social media in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ) -Most measurements were collected from the mobile operator [Spacetel (AS37424)](https://stat.ripe.net/AS37424#tabId=at-a-glance), locally known as MTN Benin, and consistently showed that the testing of the following social media sites presented HTTP failures: [facebook.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Ffacebook.com), [whatsapp.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Fwww.whatsapp.com), [telegram.org](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Ftelegram.org), [twitter.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=http:%2F%2Ftwitter.com), [instagram.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Fwww.instagram.com), [skype.com](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fwww.skype.com), [snapchat.com](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fwww.snapchat.com), [imo.im](https://explorer.ooni.io/measurement/20190428T043240Z_AS37424_Rwc66AjfwIux5AnanLtjBjlGMHduDG4KOix2OjmNrCljGMAmc7?input=http:%2F%2Fimo.im%2F), [hangouts.google.com](https://explorer.ooni.io/measurement/20190428T032552Z_AS37424_V2ifoL0WDx09o1dSdXK1qXbF3ZBlrNcF6DjERmAT3jSGRXzO63?input=https:%2F%2Fhangouts.google.com%2F), [web.wechat.com](https://explorer.ooni.io/measurement/20190428T011523Z_AS37424_3BMK29ryPLYMbZB1vnJByHdSSX9C8XGK3ckBULg7AMXMD7pPPz?input=https:%2F%2Fweb.wechat.com%2F). Youtube though remained [accessible](https://explorer.ooni.io/measurement/20190428T025254Z_AS37424_Po0R55r3LxonVWdun5wkwlpuAdDomE4qa4k2bKrgOIWurX0jmy?input=https:%2F%2Fwww.youtube.com%2F) throughout the elections. +Most measurements were collected from the mobile operator [Spacetel (AS37424)](https://stat.ripe.net/AS37424#tabId=at-a-glance), locally known as MTN Benin, and consistently showed that the testing of the following social media sites presented HTTP failures: [facebook.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Ffacebook.com), [whatsapp.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Fwww.whatsapp.com), [telegram.org](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Ftelegram.org), [twitter.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=http:%2F%2Ftwitter.com), [instagram.com](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Fwww.instagram.com), [skype.com](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fwww.skype.com), [snapchat.com](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fwww.snapchat.com), [imo.im](https://explorer.ooni.io/measurement/20190428T043240Z_AS37424_Rwc66AjfwIux5AnanLtjBjlGMHduDG4KOix2OjmNrCljGMAmc7?input=http:%2F%2Fimo.im%2F), [hangouts.google.com](https://explorer.ooni.io/measurement/20190428T032552Z_AS37424_V2ifoL0WDx09o1dSdXK1qXbF3ZBlrNcF6DjERmAT3jSGRXzO63?input=https:%2F%2Fhangouts.google.com%2F), [web.wechat.com](https://explorer.ooni.io/measurement/20190428T011523Z_AS37424_3BMK29ryPLYMbZB1vnJByHdSSX9C8XGK3ckBULg7AMXMD7pPPz?input=https:%2F%2Fweb.wechat.com%2F). Youtube though remained [accessible](https://explorer.ooni.io/measurement/20190428T025254Z_AS37424_Po0R55r3LxonVWdun5wkwlpuAdDomE4qa4k2bKrgOIWurX0jmy?input=https:%2F%2Fwww.youtube.com%2F) throughout the elections. The blocking of WhatsApp was also detected through [OONI’s WhatsApp test](https://ooni.io/nettest/whatsapp/), which attempts to perform an HTTP GET request, TCP connection, and DNS lookup to WhatsApp’s endpoints, registration service, and web version over the vantage point of the user. -The following chart illustrates the blocking of WhatsApp on MTN (AS37424) in Benin.  +The following chart illustrates the blocking of WhatsApp on MTN (AS37424) in Benin. ![](/post/2019-benin-social-media-blocking/OONI - WhatsApp blocking in Benin.png) -**Figure 2:**WhatsApp blocking in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ)  +**Figure 2:**WhatsApp blocking in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ) Both WhatsApp’s web version (web.whatsapp.com) and the registration server used by WhatsApp’s mobile app appear to have been blocked on MTN Benin by midnight, 28th April 2019 (local time). Throughout the day, all [measurements](https://explorer.ooni.org/measurement/20190428T084731Z_AS37424_PEWoAruDdsUVHj3Do5fjh2q75YtW0zQUkciidHZDsbx7QwrQlC) collected from this network consistently showed that attempts to establish TCP connections to WhatsApp’s registration service failed, while HTTP requests to web.whatsapp.com rendered HTTP failures, with connections being reset. MTN Benin though did not block access to the addresses used by the WhatsApp application, but limited the block to merely the registration service. -It’s worth noting though that WhatsApp was accessible on two other networks: [ISOCEL (AS37090)](https://explorer.ooni.org/measurement/20190428T084647Z_AS37090_ncsfkT1r5gaqm4qiaZ3utBkfeYxcibRjllOc1Jmp10jcGSLU2w) and [OPT Benin (AS28683)](https://explorer.ooni.org/measurement/20190428T100329Z_AS28683_66GIpNZ5Bn2JdWU9nHoXH7dt8w9YEmbHNRgzgdsqapPV7NXD5d), known as Benin Telecom. Some measurements collected from Benin Telecom (AS28683) suggest “endpoint blocking”, but those are false positives due to [DNS based load balancing](https://explorer.ooni.io/measurement/20190428T063548Z_AS28683_vqNse0EZfLblie2ObeJQjatNlKFOxwW4Ap8mhEQTLqKcHjfy65) (for example, `169.54.55.206` belongs to `WhatsApp Inc.`). +It’s worth noting though that WhatsApp was accessible on two other networks: [ISOCEL (AS37090)](https://explorer.ooni.org/measurement/20190428T084647Z_AS37090_ncsfkT1r5gaqm4qiaZ3utBkfeYxcibRjllOc1Jmp10jcGSLU2w) and [OPT Benin (AS28683)](https://explorer.ooni.org/measurement/20190428T100329Z_AS28683_66GIpNZ5Bn2JdWU9nHoXH7dt8w9YEmbHNRgzgdsqapPV7NXD5d), known as Benin Telecom. Some measurements collected from Benin Telecom (AS28683) suggest “endpoint blocking”, but those are false positives due to [DNS based load balancing](https://explorer.ooni.io/measurement/20190428T063548Z_AS28683_vqNse0EZfLblie2ObeJQjatNlKFOxwW4Ap8mhEQTLqKcHjfy65) (for example, `169.54.55.206` belongs to `WhatsApp Inc.`). Unlike WhatsApp, Facebook Messenger appears to have been [accessible](https://explorer.ooni.io/measurement/20190428T084745Z_AS37424_M0wsdJFR5FZEN8GuIKzPS36cYbuKa5RUyAoM7Cs3W9yhfgkdSg) on MTN Benin (AS37424) on 28th April 2019, even though [access to facebook.com was blocked](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Ffacebook.com). -The following chart shows that [Facebook Messenger was accessible in Benin](https://explorer.ooni.io/measurement/20190428T100358Z_AS28683_B7vWp01e1GAAb6ffsg2jqoObgaJeBVxyHLQ1wVmQb3KmlNtpk9) on three different networks during the elections. +The following chart shows that [Facebook Messenger was accessible in Benin](https://explorer.ooni.io/measurement/20190428T100358Z_AS28683_B7vWp01e1GAAb6ffsg2jqoObgaJeBVxyHLQ1wVmQb3KmlNtpk9) on three different networks during the elections. ![](/post/2019-benin-social-media-blocking/OONI - Facebook Messenger in Benin.png) -**Figure 3:**Facebook Messenger testing in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ)  +**Figure 3:**Facebook Messenger testing in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ) -All measurements show that TCP connections to Facebook’s endpoints succeeded (the few [DNS anomalies](https://explorer.ooni.io/measurement/20190428T233738Z_AS37424_KNTyT3H8DJXpCW4faM8PHZ6uw3dc9mogGkM6yKyt863JHbBzmo) were false positives), suggesting that Facebook Messenger worked while facebook.com was blocked. +All measurements show that TCP connections to Facebook’s endpoints succeeded (the few [DNS anomalies](https://explorer.ooni.io/measurement/20190428T233738Z_AS37424_KNTyT3H8DJXpCW4faM8PHZ6uw3dc9mogGkM6yKyt863JHbBzmo) were false positives), suggesting that Facebook Messenger worked while facebook.com was blocked. Quite similarly, measurements collected through [OONI’s Telegram test](https://ooni.io/nettest/telegram/) show that MTN Benin [blocked telegram.org](https://explorer.ooni.io/measurement/20190428T044628Z_AS37424_wmcxWVQmJR0ootqOYGor7PUXJXWEKdPr8QIgsHOd8KpxCLGzMw?input=https:%2F%2Ftelegram.org), but they [did not block access to the Telegram mobile app](https://explorer.ooni.io/measurement/20190428T084737Z_AS37424_PDJXi8Kfcx2iyYzzRX78Gv7VHMizGCLDhvkeBp30Qu2KsCtXtV). This is illustrated through the following chart, which also shows that Telegram’s web version seemed mostly accessible on other networks. ![](/post/2019-benin-social-media-blocking/telegram-blocking-benin.png) -**Figure 4:**Telegram blocking in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ)  +**Figure 4:**Telegram blocking in Benin, Open Observatory of Network Interference (OONI) measurements, Benin: [https://explorer.ooni.io/country/BJ](https://explorer.ooni.io/country/BJ) Several circumvention tool sites, such as [purevpn.fr](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fwww.purevpn.fr), [betternet.co](https://explorer.ooni.io/measurement/20190428T052542Z_AS37424_SzAPN1eapTf9KNvmS3KnmsPIYau4Nk5QDZTHr34fhNeTeqNDJH?input=https:%2F%2Fwww.betternet.co%2F), and [tigervpn.com](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fwww.tigervpn.com), presented HTTP failures. However, these failures are likely false positives, particularly given the fact that more popular circumvention tool sites, such as psiphon.ca, were [accessible](https://explorer.ooni.io/measurement/20190428T061116Z_AS37424_aHLn8cXXrPVYF4oeHd2DXZpwZACWcrhik7an5k2LaiU9OIiAuY?input=https:%2F%2Fpsiphon.ca%2F). The testing of openvpn.com presented an [anomaly](https://explorer.ooni.io/measurement/20190428T053237Z_AS37424_n2LS7jvw22xZWsiJgaV5FRdkRU8QvaxHgjT6rxMtfplUYLua5D?input=https:%2F%2Fopenvpn.net), but this was triggered by a cloudflare captcha page (i.e., the site was accessible in Benin during the elections). @@ -93,11 +93,11 @@ Interestingly, the patterns registered for Probe 11944 when it comes to tracerou ![](/post/2019-benin-social-media-blocking/ripe-atlas-1.png) ![](/post/2019-benin-social-media-blocking/ripe-atlas-2.png) -**Figure 5:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to [www.whatsapp.com](http://www.whatsapp.com), [https://atlas.ripe.net/measurements/21083876/](https://atlas.ripe.net/measurements/21083876/), April 28, 2019\. The red pattern is registered for Probe 32381 because the probe could not resolve the URL [www.whatsapp.com](http://www.whatsapp.com). +**Figure 5:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to [www.whatsapp.com](http://www.whatsapp.com), [https://atlas.ripe.net/measurements/21083876/](https://atlas.ripe.net/measurements/21083876/), April 28, 2019. The red pattern is registered for Probe 32381 because the probe could not resolve the URL [www.whatsapp.com](http://www.whatsapp.com). ![](/post/2019-benin-social-media-blocking/ripe-atlas-3.png) -**Figure 6:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online  in Benin to [www.instagram.com](http://www.instagram.com), [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21084196](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/), April 28, 2019 +**Figure 6:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to [www.instagram.com](http://www.instagram.com), [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21084196](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/), April 28, 2019 ![](/post/2019-benin-social-media-blocking/ripe-atlas-4.png) @@ -117,15 +117,15 @@ Measurements from Probe 11944, which are gathered from AS37090, are consistent w ![](/post/2019-benin-social-media-blocking/ripe-atlas-7.png) -**Figure 10:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to Google DNS (8.8.8.8), [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21083844](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/) (AS paths inferences before the shutdown), April 28, 2019 +**Figure 10:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to Google DNS (8.8.8.8), [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21083844](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/) (AS paths inferences before the shutdown), April 28, 2019 ![](/post/2019-benin-social-media-blocking/ripe-atlas-8.png) -**Figure 11:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to Google DNS (8.8.8.8), [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21083844](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/) (AS paths inferences during the shutdown), April 28, 2019 +**Figure 11:**RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to Google DNS (8.8.8.8), [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21083844](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/) (AS paths inferences during the shutdown), April 28, 2019 ![](/post/2019-benin-social-media-blocking/ripe-atlas-9.png) -**Figure 12:** RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to 9.9.9.9 (Quad9) depicting that BENIN-IX was mostly UP on election day and that the shutdown did not occur there and proving that each network implemented/suffered from the blackout differently.  [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21084227](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/), April 28, 2019 +**Figure 12:** RIPE Atlas measurement, Recurring IPv4 traceroute measurement from all probes online in Benin to 9.9.9.9 (Quad9) depicting that BENIN-IX was mostly UP on election day and that the shutdown did not occur there and proving that each network implemented/suffered from the blackout differently. [https://atlas.ripe.net/measurements/](https://atlas.ripe.net/measurements/21083876/)[21084227](https://atlas.ripe.net/measurements/21083876/)[/](https://atlas.ripe.net/measurements/21083876/), April 28, 2019 We can deduce the following from comparing these figures: @@ -143,9 +143,9 @@ IODA detected significant Internet blackouts affecting Benin on 28th and 29th Ap The [Center for Applied Internet Data Analysis (CAIDA)](https://www.caida.org/home/) runs a project called [IODA](https://ioda.caida.org/) (short for Internet Outage Detection and Analysis), which monitors the Internet, in near-realtime, to identify macroscopic Internet outages, affecting the edge of the network (i.e. significantly impacting an AS or a large fraction of a country). IODA does so using three complementary data sources: -* **Global Internet routing (BGP):** Using data from ~500 monitors participating in the RouteViews and RIPE RIS projects to establish which network blocks are reachable based on the Internet control plane. -* **Active probing:** Continuously probing a large fraction of the (routable) IPv4 address space using [a methodology developed by the University of Southern California](https://www.isi.edu/~johnh/PAPERS/Quan13c.html) to infer when a /24 block is affected by a network outage. -* **Internet Background Radiation:** Processing unsolicited traffic reaching the UCSD Network Telescope monitoring an unutilized /8 address block. +* **Global Internet routing (BGP):** Using data from ~500 monitors participating in the RouteViews and RIPE RIS projects to establish which network blocks are reachable based on the Internet control plane. +* **Active probing:** Continuously probing a large fraction of the (routable) IPv4 address space using [a methodology developed by the University of Southern California](https://www.isi.edu/~johnh/PAPERS/Quan13c.html) to infer when a /24 block is affected by a network outage. +* **Internet Background Radiation:** Processing unsolicited traffic reaching the UCSD Network Telescope monitoring an unutilized /8 address block. Data from IODA provides insight into Internet disruptions affecting entire countries, as well as the granularity required for identifying disruptions only affecting certain networks or regions within countries. @@ -167,13 +167,13 @@ Figures 14, 15, 16 and 17 below show the occurrences of Internet blackout events ![](/post/2019-benin-social-media-blocking/ioda-2.png) -**Figure 14:** Internet Outage Detection and Analysis (IODA), AS37090 +**Figure 14:** Internet Outage Detection and Analysis (IODA), AS37090 [https://ioda.caida.org/ioda/dashboard#view=inspect&entity=asn/37090&lastView=overview&from=1556385008&until=1556557928](https://ioda.caida.org/ioda/dashboard#view=inspect&entity=asn/37090&lastView=overview&from=1556385008&until=1556557928) ![](/post/2019-benin-social-media-blocking/ioda-3.png) -**Figure 15:**  Internet Outage Detection and Analysis (IODA), AS28683 +**Figure 15:** Internet Outage Detection and Analysis (IODA), AS28683 [https://ioda.caida.org/ioda/dashboard#view=inspect&entity=asn/28683&lastView=overview&from=1556385008&until=1556557928](https://ioda.caida.org/ioda/dashboard#view=inspect&entity=asn/28683&lastView=overview&from=1556385008&until=1556557928) @@ -193,10 +193,10 @@ The figures show that the blackout events began at roughly the same time (around Figures 11 and 12 show that the blackout ended in AS28683 and AS37424 before midnight UTC on 29th April 2019, whereas the blackout ended in AS37090 at around 6 AM UTC. We also observe differences in how the blackout events manifest in IODA’s data sources. AS37090 and AS37424 see a significant drop in BGP-visible /24 blocks at the beginning of the outage. However, AS37090’s visible /24 blocks curve briefly reattains prior values before dropping again. AS28683, on the other hand, experiences only a drop in its active-probing curve initially. -In summary:  +In summary: * Four different large ASes in Benin had blackouts. These blackouts were not limited to a single AS; instead, many large ASes in Benin experienced blackouts. * The blackouts begin at roughly the same time, but end at different times; it is, therefore, possible that ASes implemented them independently. * Each AS’s blackout has a different signature in IODA's data sources; for some, the blackout is visible in the BGP data source first whereas for others, the blackout is visible in the active probing data source first. -MTN Benin (AS37424) [acknowledged the Internet disruptions](https://beninwebtv.com/2019/04/coupure-dinternet-au-benin-mtn-sort-de-sa-reserve-des-remboursements-annonces/) on 28th April 2019, but declined all responsibility, promising to reimburse its clients. +MTN Benin (AS37424) [acknowledged the Internet disruptions](https://beninwebtv.com/2019/04/coupure-dinternet-au-benin-mtn-sort-de-sa-reserve-des-remboursements-annonces/) on 28th April 2019, but declined all responsibility, promising to reimburse its clients. diff --git a/content/post/2019-blocking-abortion-rights-websites-women-on-waves-web.md b/content/post/2019-blocking-abortion-rights-websites-women-on-waves-web.md index e12af6b08..b99a705b8 100644 --- a/content/post/2019-blocking-abortion-rights-websites-women-on-waves-web.md +++ b/content/post/2019-blocking-abortion-rights-websites-women-on-waves-web.md @@ -2,7 +2,7 @@ title: "On the blocking of abortion rights websites: Women on Waves & Women on Web" author: "Joana Varon (Coding Rights), Rebecca Gomperts (Women on Waves, Women on Web), Maria Xynou (OONI), Federico Ceratto (OONI), Arturo Filastò (OONI)" date: "2019-10-29" -tags: ["country-br", "country-ir", "country-tr", "country-sa", "country-kr", "censorship"] +tags: ["country-br", "country-ir", "country-tr", "country-sa", "country-kr", "censorship", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/2019-egypt-blocks-bbc-and-alhurra.ar.md b/content/post/2019-egypt-blocks-bbc-and-alhurra.ar.md index 22828ff99..367d35ef4 100644 --- a/content/post/2019-egypt-blocks-bbc-and-alhurra.ar.md +++ b/content/post/2019-egypt-blocks-bbc-and-alhurra.ar.md @@ -2,7 +2,7 @@ title: "مصر تحجب بي بي سي والحرة: التوسع في الرقابة على الإعلام وسط الاضطرابات السياسية" author: "Ramy Raoof, Mohamed El-Taher, Mohamed Tita, Arturo Filastò, Maria Xynou" date: "2019-09-26" -tags: ["egypt", "censorship", "country-eg"] +tags: ["egypt", "censorship", "country-eg", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/2019-egypt-blocks-bbc-and-alhurra.md b/content/post/2019-egypt-blocks-bbc-and-alhurra.md index 96097029d..5eb38491f 100644 --- a/content/post/2019-egypt-blocks-bbc-and-alhurra.md +++ b/content/post/2019-egypt-blocks-bbc-and-alhurra.md @@ -2,7 +2,7 @@ title: "Egypt blocks BBC and Alhurra: Expanding media censorship amid political unrest" author: "Ramy Raoof, Mohamed El-Taher, Mohamed Tita, Arturo Filastò, Maria Xynou" date: "2019-09-26" -tags: ["egypt", "censorship", "country-eg"] +tags: ["egypt", "censorship", "country-eg", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/2019-measure-blocking-lgbtqi-sites.md b/content/post/2019-measure-blocking-lgbtqi-sites.md index 6a0628924..6f63fec4e 100644 --- a/content/post/2019-measure-blocking-lgbtqi-sites.md +++ b/content/post/2019-measure-blocking-lgbtqi-sites.md @@ -2,7 +2,7 @@ title: "Call to Action: Let’s measure the blocking of LGBTQI websites around the world!" author: "Maria Xynou" date: "2019-12-04" -tags: ["lgbtqi", "censorship"] +tags: ["lgbtqi", "censorship", "theme-human_rights"] categories: ["blog"] --- diff --git a/content/post/2020-belarus-internet-outages-website-censorship/index.md b/content/post/2020-belarus-internet-outages-website-censorship/index.md index ff4e4c3bc..23f83ef8d 100644 --- a/content/post/2020-belarus-internet-outages-website-censorship/index.md +++ b/content/post/2020-belarus-internet-outages-website-censorship/index.md @@ -3,7 +3,7 @@ title: "Belarus protests: From internet outages to pervasive website censorship" description: "This report shares OONI measurements on the blocking of websites in Belarus amid protests, following the 2020 Belarusian presidential election." author: "Maria Xynou (OONI), Arturo Filastò (OONI)" date: "2020-09-15" -tags: ["belarus", "censorship", "elections", "protests", "country-by"] +tags: ["belarus", "censorship", "elections", "protests", "country-by", "theme-news_media", "theme-circumvention", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/2020-burundi-blocks-social-media-amid-election/index.md b/content/post/2020-burundi-blocks-social-media-amid-election/index.md index 656f9a944..148acb8c8 100644 --- a/content/post/2020-burundi-blocks-social-media-amid-election/index.md +++ b/content/post/2020-burundi-blocks-social-media-amid-election/index.md @@ -3,7 +3,7 @@ title: "Burundi blocks social media amid 2020 general election" description: "This report shares OONI measurements on the blocking of social media in Burundi amid its 2020 general election." author: "Maria Xynou, Arturo Filastò" date: "2020-05-21" -tags: ["burundi", "censorship", "elections", "country-bi"] +tags: ["burundi", "censorship", "elections", "country-bi", "theme-im", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/2020-engine-evaluation-spain/index.md b/content/post/2020-engine-evaluation-spain/index.md index d9f9489b7..d5ab3e627 100644 --- a/content/post/2020-engine-evaluation-spain/index.md +++ b/content/post/2020-engine-evaluation-spain/index.md @@ -3,7 +3,7 @@ title: "Evaluating OONI’s New Measurement Engine in the Context of the Blockin description: "We evaluate OONI’s new measurement Engine in the context of the blocking Women on Web in Spain" author: "Simone Basso" date: "2020-05-27" -tags: ["spain", "censorship", "sni-blocking", "country-es"] +tags: ["spain", "censorship", "sni-blocking", "country-es", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/2020-myanmar-blocks-websites-amid-covid19/index.md b/content/post/2020-myanmar-blocks-websites-amid-covid19/index.md index 11212ff4e..b44cc388d 100644 --- a/content/post/2020-myanmar-blocks-websites-amid-covid19/index.md +++ b/content/post/2020-myanmar-blocks-websites-amid-covid19/index.md @@ -3,7 +3,7 @@ title: "Myanmar blocks “fake news” websites amid COVID-19 pandemic" description: "This report shares OONI measurements on the blocking of websites in Myanmar amid the COVID-19 pandemic." author: "Phyu Phyu Kyaw, Maria Xynou, Arturo Filastò" date: "2020-05-06" -tags: ["myanmar", "censorship", "coronavirus", "country-mm"] +tags: ["myanmar", "censorship", "coronavirus", "country-mm", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/2020-tanzania-blocks-social-media-tor-election-day/index.md b/content/post/2020-tanzania-blocks-social-media-tor-election-day/index.md index 5d09e1dd0..63af448fe 100644 --- a/content/post/2020-tanzania-blocks-social-media-tor-election-day/index.md +++ b/content/post/2020-tanzania-blocks-social-media-tor-election-day/index.md @@ -3,7 +3,7 @@ title: "Tanzania blocks social media (and Tor?) on election day" description: "This report shares OONI measurements on the blocking of social media and Tor in Tanzania amid its 2020 general election." author: "Maria Xynou" date: "2020-10-28" -tags: ["tanzania", "censorship", "elections", "country-tz"] +tags: ["tanzania", "censorship", "elections", "country-tz", "theme-im", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2020-those-unspoken-thoughts-otf-fellow-report/index.md b/content/post/2020-those-unspoken-thoughts-otf-fellow-report/index.md index 43b689a35..dcec0ba31 100644 --- a/content/post/2020-those-unspoken-thoughts-otf-fellow-report/index.md +++ b/content/post/2020-those-unspoken-thoughts-otf-fellow-report/index.md @@ -3,7 +3,7 @@ title: "Those Unspoken Thoughts: A study of censorship and media freedom in Mani description: "Research report by OTF Information Controls Fellow, Chinmayi S K." author: "Chinmayi S K (OTF Information Controls Fellow)" date: "2020-06-18" -tags: ["india", "censorship", "manipur", "fellow-report", "media-freedom"] +tags: ["india", "censorship", "manipur", "fellow-report", "media-freedom", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/2020-togo-blocks-instant-messaging-apps.md b/content/post/2020-togo-blocks-instant-messaging-apps.md index 4a8015623..8ac909b87 100644 --- a/content/post/2020-togo-blocks-instant-messaging-apps.md +++ b/content/post/2020-togo-blocks-instant-messaging-apps.md @@ -2,7 +2,7 @@ title: "Togo: Instant messaging apps blocked amid 2020 presidential election" author: "Maria Xynou, Arturo Filastò" date: "2020-02-25" -tags: ["togo", "censorship", "country-tg"] +tags: ["togo", "censorship", "country-tg", "theme-im"] categories: ["report"] --- diff --git a/content/post/2021-azerbaijan/index.md b/content/post/2021-azerbaijan/index.md index adf6edd8b..3444b5928 100644 --- a/content/post/2021-azerbaijan/index.md +++ b/content/post/2021-azerbaijan/index.md @@ -2,7 +2,7 @@ title: "Media censorship in Azerbaijan through the lens of network measurement" author: "Arzu Geybullayeva (Azerbaijan Internet Watch), Maria Xynou (OONI), Arturo Filastò (OONI)" date: "2021-07-02" -tags: ["azerbaijan", "censorship", "country-az"] +tags: ["azerbaijan", "censorship", "country-az", "theme-news_media", "theme-social_media", "theme-im", "theme-circumvention"] categories: ["report"] --- @@ -13,96 +13,96 @@ categories: ["report"] ## Key Findings -As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2021-05-16&probe_cc=AZ) between January 2020 to May 2021, we found: +As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2021-05-16&probe_cc=AZ) between January 2020 to May 2021, we found: -* **Blocking of independent news media and circumvention tool websites.** Throughout the testing period, several independent news media and circumvention tool sites presented HTTP failures caused by connection timeouts. This suggests the potential use of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi) by ISPs in Azerbaijan. -* **Attempts to block Tor and Psiphon.** ISPs in Azerbaijan [attempted to block Tor](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL) and [Psiphon](https://explorer.ooni.org/measurement/20201014T041710Z_AS203622_dzuDmZBynGUPa8JvJS4Tq6xYe4VWpIHr3BAm3yQQfetESvbGRD) amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). However, both attempts appear to have been quite ineffective. -* **Temporary blocking of social media amid 2020 Nagorno-Karabakh war.** Between September 2020 to November 2020, several social media websites presented the same [HTTP failures](https://explorer.ooni.org/measurement/20201030T232015Z_webconnectivity_AZ_34876_n1_UWIBPqC9nIZq1L28?input=http%3A%2F%2Fwww.facebook.com%2F) (as news media and circumvention tool sites), while the testing of [WhatsApp](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT) and [Telegram](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp) presented signs of TLS level interference. -* **Variance of censorship across networks.** ISPs in Azerbaijan appear to be adopting similar censorship techniques. However, censorship varies from network to network, as different ISPs block different websites and apps at different moments in time. +* **Blocking of independent news media and circumvention tool websites.** Throughout the testing period, several independent news media and circumvention tool sites presented HTTP failures caused by connection timeouts. This suggests the potential use of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi) by ISPs in Azerbaijan. +* **Attempts to block Tor and Psiphon.** ISPs in Azerbaijan [attempted to block Tor](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL) and [Psiphon](https://explorer.ooni.org/measurement/20201014T041710Z_AS203622_dzuDmZBynGUPa8JvJS4Tq6xYe4VWpIHr3BAm3yQQfetESvbGRD) amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). However, both attempts appear to have been quite ineffective. +* **Temporary blocking of social media amid 2020 Nagorno-Karabakh war.** Between September 2020 to November 2020, several social media websites presented the same [HTTP failures](https://explorer.ooni.org/measurement/20201030T232015Z_webconnectivity_AZ_34876_n1_UWIBPqC9nIZq1L28?input=http%3A%2F%2Fwww.facebook.com%2F) (as news media and circumvention tool sites), while the testing of [WhatsApp](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT) and [Telegram](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp) presented signs of TLS level interference. +* **Variance of censorship across networks.** ISPs in Azerbaijan appear to be adopting similar censorship techniques. However, censorship varies from network to network, as different ISPs block different websites and apps at different moments in time. ## Introduction -In light of [reports](https://www.opendemocracy.net/en/odr/azerbaijans-blocking-of-websites/) on the blocking of websites in Azerbaijan, the [Open Observatory of Network Interference (OONI)](https://ooni.org/) and [Azerbaijan Internet Watch (AIW)](https://www.az-netwatch.org/) formed a [partnership](https://ooni.org/partners/azerbaijan-internet-watch/) to collaborate on researching internet censorship in the country. Over the past year, OONI and AIW have collaborated on [collecting](https://ooni.org/post/2020-azerbaijan-ooni-probe-testing/) and regularly [analyzing](https://www.az-netwatch.org/wp-content/uploads/2021/04/2021-04-Azerbaijan-Internet-Watch-Updates.pdf) censorship measurements from Azerbaijan, while providing timely updates through [reports](https://www.az-netwatch.org/category/technical-analysis/). +In light of [reports](https://www.opendemocracy.net/en/odr/azerbaijans-blocking-of-websites/) on the blocking of websites in Azerbaijan, the [Open Observatory of Network Interference (OONI)](https://ooni.org/) and [Azerbaijan Internet Watch (AIW)](https://www.az-netwatch.org/) formed a [partnership](https://ooni.org/partners/azerbaijan-internet-watch/) to collaborate on researching internet censorship in the country. Over the past year, OONI and AIW have collaborated on [collecting](https://ooni.org/post/2020-azerbaijan-ooni-probe-testing/) and regularly [analyzing](https://www.az-netwatch.org/wp-content/uploads/2021/04/2021-04-Azerbaijan-Internet-Watch-Updates.pdf) censorship measurements from Azerbaijan, while providing timely updates through [reports](https://www.az-netwatch.org/category/technical-analysis/). -In this report, we share findings from our analysis of [OONI network measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) between 1st January 2020 to 1st May 2021. The aim of this study is to document and increase transparency of internet censorship in Azerbaijan through the analysis of empirical network measurement data. The following sections document our methodology and present the findings from our analysis. +In this report, we share findings from our analysis of [OONI network measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) between 1st January 2020 to 1st May 2021. The aim of this study is to document and increase transparency of internet censorship in Azerbaijan through the analysis of empirical network measurement data. The following sections document our methodology and present the findings from our analysis. ## Methods -To measure internet censorship in Azerbaijan, we analyzed [OONI network measurements](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) collected from the country between 1st January 2020 to 1st May 2021. Further details are shared below. +To measure internet censorship in Azerbaijan, we analyzed [OONI network measurements](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ)collected from the country between 1st January 2020 to 1st May 2021. Further details are shared below. ### OONI network measurement -Since 2012, OONI builds [free and open source software](https://github.com/ooni/probe) (called [OONI Probe](https://ooni.org/install/)) designed to measure the [blocking of websites](https://ooni.org/nettest/web-connectivity/), instant messaging apps ([WhatsApp](https://ooni.org/nettest/whatsapp/), [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/), [Telegram](https://ooni.org/nettest/telegram/), [Signal](https://ooni.org/nettest/signal/)), circumvention tools ([Psiphon](https://ooni.org/nettest/psiphon/), [Tor](https://ooni.org/nettest/tor/), [RiseupVPN](https://ooni.org/nettest/riseupvpn/)), as well as the [speed and performance](https://ooni.org/nettest/ndt/) of tested networks. Every month, OONI Probe is run by communities in [around 200 countries](https://explorer.ooni.org/) and territories. By default, OONI Probe test results are automatically submitted to OONI servers, [processed](https://github.com/ooni/pipeline), and [openly published](https://ooni.org/data/) in near real-time. +Since 2012, OONI builds [free and open source software](https://github.com/ooni/probe) (called [OONI Probe](https://ooni.org/install/)) designed to measure the [blocking of websites](https://ooni.org/nettest/web-connectivity/), instant messaging apps ([WhatsApp](https://ooni.org/nettest/whatsapp/), [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/), [Telegram](https://ooni.org/nettest/telegram/), [Signal](https://ooni.org/nettest/signal/)), circumvention tools ([Psiphon](https://ooni.org/nettest/psiphon/), [Tor](https://ooni.org/nettest/tor/), [RiseupVPN](https://ooni.org/nettest/riseupvpn/)), as well as the [speed and performance](https://ooni.org/nettest/ndt/) of tested networks. Every month, OONI Probe is run by communities in [around 200 countries](https://explorer.ooni.org/) and territories. By default, OONI Probe test results are automatically submitted to OONI servers, [processed](https://github.com/ooni/pipeline), and [openly published](https://ooni.org/data/) in near real-time. To examine internet censorship in Azerbaijan, we: -* [Updated](https://github.com/citizenlab/test-lists/pulls?q=is%3Apr+is%3Aclosed+az) the [Citizen Lab test list for Azerbaijan](https://github.com/citizenlab/test-lists/blob/master/lists/az.csv) (which is used by OONI Probe) to include more websites for testing; -* Embedded an [“Azerbaijan” button](https://ooni.org/get-involved/run) (on the OONI website) to enable the testing of 57 Azerbaijani websites and [promoted its use](https://www.az-netwatch.org/news/measuring-website-blocks-in-azerbaijan-ooni-forensics/) for testing; -* Published a [blog post](https://ooni.org/post/2020-azerbaijan-ooni-probe-testing/) (in Azerbaijani) with relevant OONI Probe testing instructions to encourage further testing in Azerbaijan; -* Coordinated [OONI Probe](https://ooni.org/install/) testing directly with volunteers in Azerbaijan; -* Analyzed [OONI measurements](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) collected (and openly published) from Azerbaijan. +* [Updated](https://github.com/citizenlab/test-lists/pulls?q=is%3Apr+is%3Aclosed+az) the [Citizen Lab test list for Azerbaijan](https://github.com/citizenlab/test-lists/blob/master/lists/az.csv) (which is used by OONI Probe) to include more websites for testing; +* Embedded an [“Azerbaijan” button](https://ooni.org/get-involved/run) (on the OONI website) to enable the testing of 57 Azerbaijani websites and [promoted its use](https://www.az-netwatch.org/news/measuring-website-blocks-in-azerbaijan-ooni-forensics/) for testing; +* Published a [blog post](https://ooni.org/post/2020-azerbaijan-ooni-probe-testing/) (in Azerbaijani) with relevant OONI Probe testing instructions to encourage further testing in Azerbaijan; +* Coordinated [OONI Probe](https://ooni.org/install/) testing directly with volunteers in Azerbaijan; +* Analyzed [OONI measurements](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) collected (and openly published) from Azerbaijan. -As part of our [preliminary analyses](https://www.az-netwatch.org/news/how-websites-are-blocked-in-azerbaijan/), we found that a number of independent news media websites presented signs of blocking (corroborating [past reports](https://www.opendemocracy.net/en/odr/azerbaijans-blocking-of-websites/)), along with several circumvention tool websites. We therefore decided to narrow the scope of our research to primarily study the blocking of such websites, particularly since they did not present [block pages](https://ooni.org/support/glossary/#block-page) (which would have enabled us to automatically confirm their blocking), and required further ongoing testing. +As part of our [preliminary analyses](https://www.az-netwatch.org/news/how-websites-are-blocked-in-azerbaijan/), we found that a number of independent news media websites presented signs of blocking (corroborating [past reports](https://www.opendemocracy.net/en/odr/azerbaijans-blocking-of-websites/)), along with several circumvention tool websites. We therefore decided to narrow the scope of our research to primarily study the blocking of such websites, particularly since they did not present [block pages](https://ooni.org/support/glossary/#block-page) (which would have enabled us to automatically confirm their blocking), and required further ongoing testing. -To measure the blocking of websites, we analyzed measurements collected from [OONI’s Web Connectivity test](https://ooni.org/nettest/web-connectivity/), which is designed to measure the DNS, TCP/IP, and HTTP blocking of websites (included in the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists)). To measure a URL, OONI’s Web Connectivity test performs a [DNS lookup](https://ooni.org/support/glossary/#dns-lookup), attempts to establish a [TCP connection](https://ooni.org/support/glossary/#tcp) (to the list of IP addresses identified during the DNS lookup), and perform an [HTTP GET request](https://ooni.org/support/glossary/#http-request) (to retrieve the content of the tested website). These steps are automatically performed over two vantage points: the local network of the OONI Probe user, and a “control” (non-censored) network. The test results from the two networks are automatically compared and if they’re the same, the tested URL is considered accessible from the local vantage point of the user. But if the test results differ, the measurement of the tested URL is flagged as an “[anomaly](https://ooni.org/support/faq/#how-can-i-interpret-ooni-data)”, which indicates the potential presence of censorship. +To measure the blocking of websites, we analyzed measurements collected from [OONI’s Web Connectivity test](https://ooni.org/nettest/web-connectivity/), which is designed to measure the DNS, TCP/IP, and HTTP blocking of websites (included in the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists)). To measure a URL, OONI’s Web Connectivity test performs a [DNS lookup](https://ooni.org/support/glossary/#dns-lookup), attempts to establish a [TCP connection](https://ooni.org/support/glossary/#tcp) (to the list of IP addresses identified during the DNS lookup), and perform an [HTTP GET request](https://ooni.org/support/glossary/#http-request) (to retrieve the content of the tested website). These steps are automatically performed over two vantage points: the local network of the OONI Probe user, and a “control” (non-censored) network. The test results from the two networks are automatically compared and if they’re the same, the tested URL is considered accessible from the local vantage point of the user. But if the test results differ, the measurement of the tested URL is flagged as an “[anomaly](https://ooni.org/support/faq/#how-can-i-interpret-ooni-data)”, which indicates the potential presence of censorship. -OONI only automatically confirms the blocking of websites when a [block page](https://ooni.org/support/glossary/#block-page) is served. When Internet Service Providers (ISPs) serve block pages (in countries that include [Iran](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&probe_cc=IR&only=confirmed), [Russia](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&probe_cc=IN&only=confirmed), and [Italy](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&probe_cc=IT&only=confirmed), among many others), the user does not view the content of the website they’re trying to access. Instead, they view a web page (the block page) that informs them that they’re not allowed to access the intended website (and block pages often reference the law behind the censorship). There is therefore no ambiguity on whether a site is blocked or not (when a block page is served), since the ISP clearly informs the internet user that access to a specific website is censored. Moreover, block pages contain fingerprints, which OONI can add to its database in order to automatically detect and confirm cases of blocking (based on those block pages). +OONI only automatically confirms the blocking of websites when a [block page](https://ooni.org/support/glossary/#block-page) is served. When Internet Service Providers (ISPs) serve block pages (in countries that include [Iran](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&probe_cc=IR&only=confirmed), [Russia](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&probe_cc=IN&only=confirmed), and [Italy](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&probe_cc=IT&only=confirmed), among many others), the user does not view the content of the website they’re trying to access. Instead, they view a web page (the block page) that informs them that they’re not allowed to access the intended website (and block pages often reference the law behind the censorship). There is therefore no ambiguity on whether a site is blocked or not (when a block page is served), since the ISP clearly informs the internet user that access to a specific website is censored. Moreover, block pages contain fingerprints, which OONI can add to its database in order to automatically detect and confirm cases of blocking (based on those block pages). -However, many ISPs around the world -- and particularly in Azerbaijan -- adopt different censorship techniques, which are more subtle and harder to (automatically) confirm. In other cases of blocking (beyond a block page), an internet user may see an error, which may not be clear if the website is intentionally censored, if the website is temporarily down, if there’s a case of server-side blocking (where the website owner blocks IP addresses from a specific country, rather than an ISP implementing government-mandated censorship), or if the user is unable to access the website due to transient network failures. There are a number of reasons why a website may be inaccessible, many of which may have nothing to do with intentional blocking, and can potentially lead to [false positives](https://ooni.org/support/faq/#what-are-false-positives) in OONI measurements. This is why OONI flags all other cases (beyond block pages) as “[anomalies](https://ooni.org/support/faq/#how-can-i-interpret-ooni-data)”: to signal a potential case of blocking, while acknowledging the risk of false positives.   +However, many ISPs around the world -- and particularly in Azerbaijan -- adopt different censorship techniques, which are more subtle and harder to (automatically) confirm. In other cases of blocking (beyond a block page), an internet user may see an error, which may not be clear if the website is intentionally censored, if the website is temporarily down, if there’s a case of server-side blocking (where the website owner blocks IP addresses from a specific country, rather than an ISP implementing government-mandated censorship), or if the user is unable to access the website due to transient network failures. There are a number of reasons why a website may be inaccessible, many of which may have nothing to do with intentional blocking, and can potentially lead to [false positives](https://ooni.org/support/faq/#what-are-false-positives) in OONI measurements. This is why OONI flags all other cases (beyond block pages) as “[anomalies](https://ooni.org/support/faq/#how-can-i-interpret-ooni-data)”: to signal a potential case of blocking, while acknowledging the risk of false positives. To evaluate whether “anomalous” measurements from Azerbaijan are indicators of blocking events, we analyzed relevant measurements in aggregate to examine whether the same type of anomalies (“HTTP failures”) were persistent (for the same URLs) on the same local networks over time. In other words, if a tested URL presents the same anomalies (in terms of network measurement data results) every time (or most times) it was tested on a specific network in Azerbaijan over a long period of time, that serves as a strong indicator that that website was blocked on that network. If, however, a tested URL presented a small ratio of anomalous measurements (in comparison to the overall measurements that indicated accessibility), it was excluded from our findings (since those few anomalous measurements were likely false positives). -As several circumvention tool websites consistently presented anomalies throughout the testing period, we also analyzed OONI measurements collected from OONI’s [Tor](https://ooni.org/nettest/tor/) and [Psiphon](https://ooni.org/nettest/psiphon/) tests. OONI’s [Tor](https://ooni.org/nettest/tor/) test is designed to measure whether [Tor](https://www.torproject.org/) (free software for online anonymity, privacy, and censorship circumvention) works from a tested network. To this end, the test measures the reachability of selected Tor directory authorities and [bridges](https://bridges.torproject.org/); if it succeeds, Tor may work on the tested network, but if it fails, access to Tor may be blocked. OONI’s [Psiphon](https://ooni.org/nettest/psiphon/) test, on the other hand, attempts to bootstrap the Psiphon tunnel and use it to fetch a web page from the internet. If this process is successful, the Psiphon VPN app may work on the tested network. But if the test is unable to fetch a web page (using the Psiphon tunnel), or is unable to bootstrap Psiphon at all, access to Psiphon may be blocked on the local network. +As several circumvention tool websites consistently presented anomalies throughout the testing period, we also analyzed OONI measurements collected from OONI’s [Tor](https://ooni.org/nettest/tor/) and [Psiphon](https://ooni.org/nettest/psiphon/) tests. OONI’s [Tor](https://ooni.org/nettest/tor/) test is designed to measure whether [Tor](https://www.torproject.org/) (free software for online anonymity, privacy, and censorship circumvention) works from a tested network. To this end, the test measures the reachability of selected Tor directory authorities and [bridges](https://bridges.torproject.org/); if it succeeds, Tor may work on the tested network, but if it fails, access to Tor may be blocked. OONI’s [Psiphon](https://ooni.org/nettest/psiphon/) test, on the other hand, attempts to bootstrap the Psiphon tunnel and use it to fetch a web page from the internet. If this process is successful, the Psiphon VPN app may work on the tested network. But if the test is unable to fetch a web page (using the Psiphon tunnel), or is unable to bootstrap Psiphon at all, access to Psiphon may be blocked on the local network. -Amid the Nagorno-Karabakh war (between September to November in 2020), access to online social media was [reportedly blocked](https://www.washingtonpost.com/politics/2020/12/04/while-armenia-azerbaijan-fought-over-nagorno-karabakh-their-citizens-battled-social-media/) in Azerbaijan. We therefore analyzed OONI [Web Connectivity](https://ooni.org/nettest/web-connectivity/) measurements (discussed previously) pertaining to the testing of social media websites, as well as measurements collected from OONI’s [WhatsApp](https://ooni.org/nettest/whatsapp/), [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/), and [Telegram](https://ooni.org/nettest/telegram/) tests.   +Amid the Nagorno-Karabakh war (between September to November in 2020), access to online social media was [reportedly blocked](https://www.washingtonpost.com/politics/2020/12/04/while-armenia-azerbaijan-fought-over-nagorno-karabakh-their-citizens-battled-social-media/) in Azerbaijan. We therefore analyzed OONI [Web Connectivity](https://ooni.org/nettest/web-connectivity/) measurements (discussed previously) pertaining to the testing of social media websites, as well as measurements collected from OONI’s [WhatsApp](https://ooni.org/nettest/whatsapp/), [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/), and [Telegram](https://ooni.org/nettest/telegram/) tests. -OONI’s [WhatsApp](https://ooni.org/nettest/whatsapp/) test is designed to examine the reachability of both WhatsApp’s app and the WhatsApp web version (web.whatsapp.com) within a tested network. To measure the WhatsApp app, this test attempts to perform TCP connections to WhatsApp’s endpoints and registration service, DNS lookups (to resolve IP addresses allocated to WhatsApp), and HTTP requests to WhatsApp’s registration service. If any of these steps fail, access to the WhatsApp app is likely blocked. To measure the WhatsApp web interface, the test attempts to perform a DNS lookup, TCP connection, and HTTP request to web.whatsapp.com. If these attempts fail or return inconsistent responses, access to web.whatsapp.com is likely blocked. +OONI’s [WhatsApp](https://ooni.org/nettest/whatsapp/) test is designed to examine the reachability of both WhatsApp’s app and the WhatsApp web version (web.whatsapp.com) within a tested network. To measure the WhatsApp app, this test attempts to perform TCP connections to WhatsApp’s endpoints and registration service, DNS lookups (to resolve IP addresses allocated to WhatsApp), and HTTP requests to WhatsApp’s registration service. If any of these steps fail, access to the WhatsApp app is likely blocked. To measure the WhatsApp web interface, the test attempts to perform a DNS lookup, TCP connection, and HTTP request to web.whatsapp.com. If these attempts fail or return inconsistent responses, access to web.whatsapp.com is likely blocked. -OONI’s [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/) test attempts to perform DNS lookups (to domains associated with Facebook) and TCP connections to Facebook’s endpoints. If the DNS lookups do not resolve to IP addresses allocated to Facebook, and/or TCP connections to Facebook endpoints fail, access to Facebook Messenger is likely blocked. OONI’s [Telegram](https://ooni.org/nettest/telegram/) test attempts to perform TCP connections and HTTP POST requests to Telegram access points, as well as HTTP(S) GET requests to web.telegram.org. If these attempts fail or return inconsistent responses, access to Telegram’s app and/or web.telegram.org is likely blocked. +OONI’s [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/) test attempts to perform DNS lookups (to domains associated with Facebook) and TCP connections to Facebook’s endpoints. If the DNS lookups do not resolve to IP addresses allocated to Facebook, and/or TCP connections to Facebook endpoints fail, access to Facebook Messenger is likely blocked. OONI’s [Telegram](https://ooni.org/nettest/telegram/) test attempts to perform TCP connections and HTTP POST requests to Telegram access points, as well as HTTP(S) GET requests to web.telegram.org. If these attempts fail or return inconsistent responses, access to Telegram’s app and/or web.telegram.org is likely blocked. We analyzed all OONI measurements (from the aforementioned tests) collected from multiple local networks in Azerbaijan between 1st January 2020 to 1st May 2021. We discuss methodological limitations and present our findings in the following sections. #### Acknowledgement of limitations -[OONI measurement coverage from Azerbaijan](https://explorer.ooni.org/country/AZ) has been rather limited over the years (in comparison to other countries), limiting our findings and confidence in confirming censorship events. OONI data (and our censorship measurement findings) depend entirely on volunteers running [OONI Probe](https://ooni.org/install/) locally in Azerbaijan, which can be challenging in light of the political environment, [potential risks](https://ooni.org/about/risks/) involved with running OONI Probe, and the relatively limited digital rights community engagement opportunities in Azerbaijan. Despite our [best efforts](https://ooni.org/post/2020-azerbaijan-ooni-probe-testing/) to encourage more OONI Probe testing, OONI measurement coverage remained relatively low throughout the testing period. +[OONI measurement coverage from Azerbaijan](https://explorer.ooni.org/country/AZ) has been rather limited over the years (in comparison to other countries), limiting our findings and confidence in confirming censorship events. OONI data (and our censorship measurement findings) depend entirely on volunteers running [OONI Probe](https://ooni.org/install/) locally in Azerbaijan, which can be challenging in light of the political environment, [potential risks](https://ooni.org/about/risks/) involved with running OONI Probe, and the relatively limited digital rights community engagement opportunities in Azerbaijan. Despite our [best efforts](https://ooni.org/post/2020-azerbaijan-ooni-probe-testing/) to encourage more OONI Probe testing, OONI measurement coverage remained relatively low throughout the testing period. But starting from 24th December 2020, OONI Probe testing coverage in Azerbaijan increased significantly. The following chart illustrates that most domains from the [testing lists](https://ooni.org/support/faq/#which-websites-will-i-test-for-censorship-with-ooni-probe) were tested at least once between December 2020 to February 2021, while in many cases, around 1,400 domains were tested per day in Azerbaijan. ![](images/image11.png) -Chart: Distinct count of tested domains in Azerbaijan between October 2020 to February 2021. +**Chart:** Distinct count of tested domains in Azerbaijan between October 2020 to February 2021. Moreover, the overall volume of OONI measurements gathered every day from Azerbaijan (between December 2020 to February 2021) was quite stable, as illustrated through the following chart. ![](images/image6.png) -Chart: Number of OONI measurements collected per day (on the testing of websites) from Azerbaijan between October 2020 to February 2021. +**Chart:** Number of OONI measurements collected per day (on the testing of websites) from Azerbaijan between October 2020 to February 2021. It’s worth highlighting that, on average, more than 1,000 measurements were collected per day between January 2021 to February 2021. On several days (during this time frame), around 3,000 measurements were collected from Azerbaijan per day. Apart from the overall measurement coverage, other limitations to this study are associated with per network measurement coverage and the types of domains tested. Overall, this study includes analysis of OONI measurements collected from 31 unique AS networks, but that does not include all AS networks in Azerbaijan. It is therefore possible that we may have missed censorship events on networks that weren’t tested in Azerbaijan during the analysis period. The relatively scarce and sporadic testing across AS networks also means that we may have missed potential temporary blocking events on networks that generally received measurement coverage. -While we tried to ensure that the Citizen Lab test lists include both [internationally-relevant websites](https://github.com/citizenlab/test-lists/blob/master/lists/global.csv) and [websites relevant to Azerbaijan](https://github.com/citizenlab/test-lists/blob/master/lists/az.csv) (including those that were previously reported as blocked by the media), the overall number of URLs tested was rather limited (i.e. we did not test every website available on the internet). Therefore, websites which may be blocked in Azerbaijan, but which were not included in the relevant [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists), were not tested and are not included in our findings. Moreover, not all websites were tested on all networks (that received general testing coverage), nor were they all tested consistently over time, which means that we may have missed the blocking of some websites on some networks. +While we tried to ensure that the Citizen Lab test lists include both [internationally-relevant websites](https://github.com/citizenlab/test-lists/blob/master/lists/global.csv) and [websites relevant to Azerbaijan](https://github.com/citizenlab/test-lists/blob/master/lists/az.csv) (including those that were previously reported as blocked by the media), the overall number of URLs tested was rather limited (i.e. we did not test every website available on the internet). Therefore, websites which may be blocked in Azerbaijan, but which were not included in the relevant [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists), were not tested and are not included in our findings. Moreover, not all websites were tested on all networks (that received general testing coverage), nor were they all tested consistently over time, which means that we may have missed the blocking of some websites on some networks. -Finally, our app testing was limited to OONI’s [WhatsApp](https://ooni.org/nettest/whatsapp/), [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/), [Telegram](https://ooni.org/nettest/telegram/), [Psiphon](https://ooni.org/nettest/psiphon/), and [Tor](https://ooni.org/nettest/tor/) tests, which means that other apps (which may have been blocked during our analysis period) were not tested and therefore excluded from the findings of this study.   +Finally, our app testing was limited to OONI’s [WhatsApp](https://ooni.org/nettest/whatsapp/), [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/), [Telegram](https://ooni.org/nettest/telegram/), [Psiphon](https://ooni.org/nettest/psiphon/), and [Tor](https://ooni.org/nettest/tor/) tests, which means that other apps (which may have been blocked during our analysis period) were not tested and therefore excluded from the findings of this study. ## Background When the internet arrived in Azerbaijan in 1996, little did its users know that in a little over than decade, it would change the landscape of political and civic activism in the country. Similarly, the likelihood of the internet and its services, used for accessing information, raising awareness, campaigning, news sharing, and political activism, was not on the government agenda. Until it became evident, that this may as well be the case. -The most recent [available data](https://theinclusiveinternet.eiu.com/explore/countries/performance/availability/usage/internet-users?highlighted=AZ.;) shows that the internet penetration rate in Azerbaijan is at 80 percent. Having said that, internet quality and access remains poor and relatively inaccessible to the rural communities. +The most recent [available data](https://theinclusiveinternet.eiu.com/explore/countries/performance/availability/usage/internet-users?highlighted=AZ.;) shows that the internet penetration rate in Azerbaijan is at 80 percent. Having said that, internet quality and access remains poor and relatively inaccessible to the rural communities. -Despite the adoption of government plans and programs aimed at promoting internet development and freedoms, the government [continues to fully control](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) the internet landscape, including internet infrastructure, access, and service quality in an environment [marred by human rights violations](https://www.amnesty.org/en/countries/europe-and-central-asia/azerbaijan/) and the state-of-the-art surveillance technology [utilized](https://www.academia.edu/37482634/Match_made_in_heaven_Authoritarian_states_and_digital_surveillance_Case_study_from_Azerbaijan) to [silence dissent](https://www.codastory.com/authoritarian-tech/surveillance-and-internet-disruption-in-baku/). +Despite the adoption of government plans and programs aimed at promoting internet development and freedoms, the government [continues to fully control](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) the internet landscape, including internet infrastructure, access, and service quality in an environment [marred by human rights violations](https://www.amnesty.org/en/countries/europe-and-central-asia/azerbaijan/) and the state-of-the-art surveillance technology [utilized](https://www.academia.edu/37482634/Match_made_in_heaven_Authoritarian_states_and_digital_surveillance_Case_study_from_Azerbaijan) to [silence dissent](https://www.codastory.com/authoritarian-tech/surveillance-and-internet-disruption-in-baku/). -The Ministry of Transportation, Communication and High Technologies (MTCHT) remains the main internet regulator in the country, although the country’s Strategic Road Map for Telecommunications and Information Technologies for 2016-2020 [called for](https://monitoring.az/assets/upload/files/6683729684f8895c1668803607932190.pdf) the establishment of an independent regulatory body. Thus, there is currently a lack of independent oversight of the ICT sector. At the same time, [national legislation](http://www.e-qanun.az/framework/32626) requires telecommunications operators and providers to cooperate with law enforcement authorities on the implementation of surveillance schemes, in particular by installing special equipment that grants officials access to confidential information about their customers, in violation of international data protection standards. +The Ministry of Transportation, Communication and High Technologies (MTCHT) remains the main internet regulator in the country, although the country’s Strategic Road Map for Telecommunications and Information Technologies for 2016-2020 [called for](https://monitoring.az/assets/upload/files/6683729684f8895c1668803607932190.pdf) the establishment of an independent regulatory body. Thus, there is currently a lack of independent oversight of the ICT sector. At the same time, [national legislation](http://www.e-qanun.az/framework/32626) requires telecommunications operators and providers to cooperate with law enforcement authorities on the implementation of surveillance schemes, in particular by installing special equipment that grants officials access to confidential information about their customers, in violation of international data protection standards. -The licensing process is regulated by the 2016 Law on Licenses and Permits and the  Resolution No. 174 on "additional conditions required for the issuance of special permits (licenses) depending on the nature of the activity", which was adopted by the Cabinet of Ministers on 7 November 2002. This resolution requires a written affirmation of their readiness to install special equipment needed for law enforcement authorities to obtain information from telecommunications networks for reasons of national security. The license fee depends on the nature of the telecommunications services provided. According to Resolution No. 427 on the “Rules of registration of operators and providers of Internet telecommunication services”, which was approved by the Cabinet of Ministers on 12 October 2017, ISPs are required to present a guarantee that they will install special equipment needed to access information during search operations conducted by authorities. +The licensing process is regulated by the 2016 Law on Licenses and Permits and the Resolution No. 174 on "additional conditions required for the issuance of special permits (licenses) depending on the nature of the activity", which was adopted by the Cabinet of Ministers on 7 November 2002. This resolution requires a written affirmation of their readiness to install special equipment needed for law enforcement authorities to obtain information from telecommunications networks for reasons of national security. The license fee depends on the nature of the telecommunications services provided. According to Resolution No. 427 on the “Rules of registration of operators and providers of Internet telecommunication services”, which was approved by the Cabinet of Ministers on 12 October 2017, ISPs are required to present a guarantee that they will install special equipment needed to access information during search operations conducted by authorities. ### Network landscape and internet penetration In addition to underdeveloped infrastructure, and widespread disparities in internet access and quality, the international connectivity of the broadband network is in need of an upgrade and expansion. Limited competition for fixed broadband infrastructure has led to complacency and slow internet speeds. -There are two companies in Azerbaijan - Delta Telekom and AzerTelekom - holding monopoly on the provision of international internet access and selling such services to other national Internet Service Providers (ISPs).  Delta is state owned while AzerTelekom is controlled by the state run Azerfon mobile company.  The rest of the ICT market is also dominated by state-owned companies. +There are two companies in Azerbaijan - Delta Telekom and AzerTelekom - holding monopoly on the provision of international internet access and selling such services to other national Internet Service Providers (ISPs). Delta is state owned while AzerTelekom is controlled by the state run Azerfon mobile company. The rest of the ICT market is also dominated by state-owned companies. ![](images/image4.png) @@ -116,19 +116,19 @@ Infographic 2 developed by [MikroskopMedia](https://mikroskopmedia.com/) The cost of the internet packages offered by ISPs ranges between 15AZN (10.5 USD) and 80 AZN (47 USD) per month. In comparison, the average monthly salary in Azerbaijan is estimated by independent experts to be about 340 AZN (200 USD), making the costs of internet access is prohibitively expensive for many in the country (see Infographic 2). -According to the 2020 Inclusive Internet Index, the [monthly cost of a fixed internet](https://theinclusiveinternet.eiu.com/explore/countries/performance/affordability/price/fixed-line-monthly-broadband-cost?highlighted=AZ) connection in Azerbaijan is 2 percent of the gross national income (GNI) per capita, while the [cost of a mobile data plan](https://theinclusiveinternet.eiu.com/explore/countries/performance/affordability/price/mobile-phone-cost-prepaid-tariff?highlighted=AZ) offering 1 GB per month is 1 percent of the GNI per capita, leaving Azerbaijan behind countries like Ukraine, Turkey and Russia. +According to the 2020 Inclusive Internet Index, the [monthly cost of a fixed internet](https://theinclusiveinternet.eiu.com/explore/countries/performance/affordability/price/fixed-line-monthly-broadband-cost?highlighted=AZ) connection in Azerbaijan is 2 percent of the gross national income (GNI) per capita, while the [cost of a mobile data plan](https://theinclusiveinternet.eiu.com/explore/countries/performance/affordability/price/mobile-phone-cost-prepaid-tariff?highlighted=AZ) offering 1 GB per month is 1 percent of the GNI per capita, leaving Azerbaijan behind countries like Ukraine, Turkey and Russia. ### Legal environment -The government of Azerbaijan relies on the [Law on Telecommunications](http://www.e-qanun.az/framework/10663) and legal norms to control the ISPs and mobile operators. Although the law sets out as the basic principle the separation of state regulatory functions from telecommunications, in reality, the law grants the government broad powers with respect to the licensing, and certification of telecommunication activities; the application of tariffs for the use of telecommunication services and radio frequencies; and the regulation of competition, antimonopoly activity and other aspects of telecommunications.   +The government of Azerbaijan relies on the [Law on Telecommunications](http://www.e-qanun.az/framework/10663) and legal norms to control the ISPs and mobile operators. Although the law sets out as the basic principle the separation of state regulatory functions from telecommunications, in reality, the law grants the government broad powers with respect to the licensing, and certification of telecommunication activities; the application of tariffs for the use of telecommunication services and radio frequencies; and the regulation of competition, antimonopoly activity and other aspects of telecommunications. -Recent decrees and legal amendments further signal the state’s approach to internet freedoms and independent media environment. On January 12, President Ilham Aliyev signed a decree “on deepening media reforms in the Republic of Azerbaijan.” As a result, the newly [established](https://www.azernews.az/nation/174953.html) Azerbaijani Agency for Media Development will replace the State Support Fund for Mass Media Development and will have all the rights of the former institution. In tandem, new media law is also being [drafted](https://apa.az/en/media-news/Azerbaijan-to-prepare-law-on-Media-340019) by the Administration of the President for the President’s review in two months. According to its charter, the agency can take measures to protect state and commercial secrets. In case of non-compliance with the information published in the online media within the requirements provided by law, the agency can contact the relevant authorities in order to take measures in this regard. +Recent decrees and legal amendments further signal the state’s approach to internet freedoms and independent media environment. On January 12, President Ilham Aliyev signed a decree “on deepening media reforms in the Republic of Azerbaijan.” As a result, the newly [established](https://www.azernews.az/nation/174953.html) Azerbaijani Agency for Media Development will replace the State Support Fund for Mass Media Development and will have all the rights of the former institution. In tandem, new media law is also being [drafted](https://apa.az/en/media-news/Azerbaijan-to-prepare-law-on-Media-340019) by the Administration of the President for the President’s review in two months. According to its charter, the agency can take measures to protect state and commercial secrets. In case of non-compliance with the information published in the online media within the requirements provided by law, the agency can contact the relevant authorities in order to take measures in this regard. It also has the authority to take measures in accordance with the Code of Administrative Offenses in case of detecting signs of an administrative violation in the field of print and online media, and in case of detection of signs of a crime – to provide information to the appropriate authority the powers are similar to the National Council on Television and Radio which can and has in the past [deprived](https://www.turan.az/ext/news/2021/1/free/politics%20news/en/337.htm) radio and television companies of air hours. It will also be accountable to the head of state. -The agency’s governing bodies – a Supervisory Board of six members – and the executive director, are appointed by the head of state. Ahmed Ismayilov, is the executive director of the new fund. In April 2020, he was [appointed](http://contact.az/ext/news/2021/1/free/politics%20news/en/311.htm) the executive director of the now defunct Media Development Support Fund. Ismayilov, 40, is a lawyer by education. Previously he has worked in various government institutions, including the Heydar Aliyev Foundation, managed by the first lady and the first vice president Mehriban Aliyeva. He is a member of the ruling Yeni Azerbaijan Party. +The agency’s governing bodies – a Supervisory Board of six members – and the executive director, are appointed by the head of state. Ahmed Ismayilov, is the executive director of the new fund. In April 2020, he was [appointed](http://contact.az/ext/news/2021/1/free/politics%20news/en/311.htm) the executive director of the now defunct Media Development Support Fund. Ismayilov, 40, is a lawyer by education. Previously he has worked in various government institutions, including the Heydar Aliyev Foundation, managed by the first lady and the first vice president Mehriban Aliyeva. He is a member of the ruling Yeni Azerbaijan Party. Previously, the central executive body supervising the media in Azerbaijan was abolished when the country joined the Council of Europe in 2001. @@ -136,13 +136,13 @@ In his oOp-ed, the director of Turan News Agency, Mehman Aliyev [wrote](https:// In the meantime, notes Aliyev, while the new fund’s focus is on technological aspects of media development there is no mention of protection of free press whatsoever. Lack of avenues for independent media in the country and impunity is the challenge, not the lack of technological equipment, notesequipment notes Aliyev. -Alasgar Mammadli, the media law expert, [criticized](https://www.turan.az/ext/news/2021/1/free/politics%20news/en/333.htm) the new agency’s broad, vaguely defined legal powers and the absence of any wider preliminary discussions in the society ahead of its approval. +Alasgar Mammadli, the media law expert, [criticized](https://www.turan.az/ext/news/2021/1/free/politics%20news/en/333.htm) the new agency’s broad, vaguely defined legal powers and the absence of any wider preliminary discussions in the society ahead of its approval. Six months down the line, the draft media law, is finally set for review, albeit much to the disappointment of freedom of the media advocates and media practitioners in Azerbaijan. -According to Azadliq Radio report, the new law, entails licensing the Internet television and radio broadcasting. The [proposal](https://www.azadliq.org/a/internet-tv-lisenziya/31313244.html) spearheaded by the National Television and Radio Council (NTRC) was announced on June 17. +According to Azadliq Radio report, the new law, entails licensing the Internet television and radio broadcasting. The [proposal](https://www.azadliq.org/a/internet-tv-lisenziya/31313244.html) spearheaded by the National Television and Radio Council (NTRC) was announced on June 17. -Specifically the draft law states that: +Specifically the draft law states that: 1) the online channel must have its own website and broadcast from this site; @@ -152,44 +152,44 @@ In addition, the Agency for the Development of Mass Media would register online Several independent experts, said if true, the new bill and specifically the proposal about licensing, violate Article 10 of the European Convention on Human Rights and norms enshrined in Azerbaijan’s Constitution. -[Addressing](https://www.azadliq.org/a/internet-tv-lisenziya/31313244.html) the controversial new bill, a media law expert, Alasgar Mammadli, said in addition to contradicting [Article 10 of the Convention](https://www.echr.coe.int/documents/convention_eng.pdf) the license requirement can only be applied to broadcasters using frequency transmissions which is not the case for Internet television. In another [interview](https://www.azadliq.org/a/internet-tv-lisenziya/31313244.html), Mammadli said, *“Only during the broadcast, there should be compliance with the general law, which is currently regulated by the Law on Mass Media, Criminal Law, and other laws. There are no gaps, and there are even unnecessary regulations (restrictions).”* +[Addressing](https://www.azadliq.org/a/internet-tv-lisenziya/31313244.html) the controversial new bill, a media law expert, Alasgar Mammadli, said in addition to contradicting [Article 10 of the Convention](https://www.echr.coe.int/documents/convention_eng.pdf) the license requirement can only be applied to broadcasters using frequency transmissions which is not the case for Internet television. In another [interview](https://www.azadliq.org/a/internet-tv-lisenziya/31313244.html), Mammadli said, *“Only during the broadcast, there should be compliance with the general law, which is currently regulated by the Law on Mass Media, Criminal Law, and other laws. There are no gaps, and there are even unnecessary regulations (restrictions).”* -Another legal expert, Khaled Aghaliyev, evaluating the bill in a post on social media platform Facebook [said](https://www.facebook.com/khaled.aqaliyev/posts/10225431628275433), *“It was clear that the government, which promised progressive reforms in the legal regulation of the media, worked harder than ever on reactionary regulatory mechanisms.”* Aghaliyev said, in all likelihood, the lawyers working on “progressive regulations” took it upon themselves to interpret one specific sentence of Article 10 word for word. That sentence, notes Aghaliyev, says Aghaliyev says, *“This Article shall not prevent States from requiring the licensing of broadcasting, television or cinema enterprises.” “But they [lawyers] thought wrong. The mentioning of that licensing applies only to traditional television, and radio. Therefore, the part of the new bill that we know of, is reactionary, binding freedom of expression. It does not comply with our constitutional norms or the European Convention.”* +Another legal expert, Khaled Aghaliyev, evaluating the bill in a post on social media platform Facebook [said](https://www.facebook.com/khaled.aqaliyev/posts/10225431628275433), *“It was clear that the government, which promised progressive reforms in the legal regulation of the media, worked harder than ever on reactionary regulatory mechanisms.”* Aghaliyev said, in all likelihood, the lawyers working on “progressive regulations” took it upon themselves to interpret one specific sentence of Article 10 word for word. That sentence, notes Aghaliyev, says Aghaliyev says, *“This Article shall not prevent States from requiring the licensing of broadcasting, television or cinema enterprises.” “But they [lawyers] thought wrong. The mentioning of that licensing applies only to traditional television, and radio. Therefore, the part of the new bill that we know of, is reactionary, binding freedom of expression. It does not comply with our constitutional norms or the European Convention.”* -In March 2020 of last, members of Azerbaijan’s National Parliament approved proposed amendments to the [Law on Information, Informatisation and Protection of Information](http://www.e-qanun.az/framework/3525) during the first reading. +In March 2020 of last, members of Azerbaijan’s National Parliament approved proposed amendments to the [Law on Information, Informatisation and Protection of Information](http://www.e-qanun.az/framework/3525) during the first reading. -Special clauses A special clause on “information-telecommunication network”  and “information-telecommunication network users” were [added](http://www.meclis.gov.az/?/az/law/2380/1) to article 13.2. of the law. +Special clauses A special clause on “information-telecommunication network” and “information-telecommunication network users” were [added](http://www.meclis.gov.az/?/az/law/2380/1) to article 13.2. of the law. -While there is no definition of what the “information-telecommunication network [and its users]” clause actually means, [some](https://www.facebook.com/elesger/posts/10158086752689738) media experts and journalists suggested this referred to social media platforms and their users. In Azerbaijan, the Ministry of Transportation, Communication and High Technologies already [holds broad powers](https://www.refworld.org/docid/5a547d663.html) to block websites, without a court order. If these recent suggestions to the law are approved in the final reading, it would further deteriorate freedom of speech online as social media users, posting content the Ministry may deem as misinformation, may be arrested and face charges. +While there is no definition of what the “information-telecommunication network [and its users]” clause actually means, [some](https://www.facebook.com/elesger/posts/10158086752689738) media experts and journalists suggested this referred to social media platforms and their users. In Azerbaijan, the Ministry of Transportation, Communication and High Technologies already [holds broad powers](https://www.refworld.org/docid/5a547d663.html) to block websites, without a court order. If these recent suggestions to the law are approved in the final reading, it would further deteriorate freedom of speech online as social media users, posting content the Ministry may deem as misinformation, may be arrested and face charges. -One parliament member, Ganira Pashayeva, even [suggested](https://www.azadliq.org/a/milli-meclis-sosial-mediada-koronavirus/30495529.html) setting up a special unit that would monitor social media platforms, and hold those spreading rumors accountable. +One parliament member, Ganira Pashayeva, even [suggested](https://www.azadliq.org/a/milli-meclis-sosial-mediada-koronavirus/30495529.html) setting up a special unit that would monitor social media platforms, and hold those spreading rumors accountable. -The controversial law on Information, informatisation and access to information was first adopted in 1998. On March 10, 2017, a [series of restrictive amendments](http://www.e-qanun.az/framework/3525) were added to the law, converting the law from a technical regulation into content regulation: +The controversial law on Information, informatisation and access to information was first adopted in 1998. On March 10, 2017, a [series of restrictive amendments](http://www.e-qanun.az/framework/3525) were added to the law, converting the law from a technical regulation into content regulation: -*  **Article 13.1.3.** Creates conditions for the regulation of the domain names without the participation of relevant the parties from the internet community, but by the relevant Ministry, which contradicts international norms, including [ICANN recommendations](https://www.icann.org/resources/pages/policy-2012-02-25-en) in this regard; -* **Article 13.2.3.** All legal and ethical issues previously existing in various laws have been listed as prohibited information and it has been stressed that their dissemination is prohibited; -* **Article 13.2.4.** When the owner of the internet information resource and its domain name posts the information, dissemination of which is prohibited or receives an application about that piece of shared information, it guarantees the removal of such information from the information resource; -* **Article 13.2.5.** When a hosting provider reveals in its information systems some information, dissemination of which in internet information resources is prohibited or receives information about it, it should undertake immediate measures for its removal by the owner of the information resource; -* **Article 13.3.3.** In cases of existence of real threat for the lawful interests of the state and society or in urgent cases when there is a risk for life or health of people, the access to internet information resources is temporarily restricted directly by the Ministry of Transport, Communications and High Technologies [restriction is applied without a court order. Although an application is made to the court, the decision to close down the online information source remains in force until the court handles the case or the decision is annulled.]; -* **Article 13.3.6.** Describes the List of information resources that are “blocked” which is curated and maintained by the Ministry [to this day, no such resource is publicly available however, AIW has a list of online resources that are regularly measured for blocking with [OONI Probe](https://ooni.org/install/) monitored relying on OONI for blocking]. Independent legal experts [believe](https://rm.coe.int/azerbaijan-analysis-of-legislation-on-access-to-information-december-2/16808ae03c) this kind of authority is restrictive in nature, particularly since it forces Internet Service Providers (ISPs) in the country to block access to specific online resources. +* **Article 13.1.3.** Creates conditions for the regulation of the domain names without the participation of relevant the parties from the internet community, but by the relevant Ministry, which contradicts international norms, including [ICANN recommendations](https://www.icann.org/resources/pages/policy-2012-02-25-en) in this regard; +* **Article 13.2.3.** All legal and ethical issues previously existing in various laws have been listed as prohibited information and it has been stressed that their dissemination is prohibited; +* **Article 13.2.4.** When the owner of the internet information resource and its domain name posts the information, dissemination of which is prohibited or receives an application about that piece of shared information, it guarantees the removal of such information from the information resource; +* **Article 13.2.5.** When a hosting provider reveals in its information systems some information, dissemination of which in internet information resources is prohibited or receives information about it, it should undertake immediate measures for its removal by the owner of the information resource; +* **Article 13.3.3.** In cases of existence of real threat for the lawful interests of the state and society or in urgent cases when there is a risk for life or health of people, the access to internet information resources is temporarily restricted directly by the Ministry of Transport, Communications and High Technologies [restriction is applied without a court order. Although an application is made to the court, the decision to close down the online information source remains in force until the court handles the case or the decision is annulled.]; +* **Article 13.3.6.** Describes the List of information resources that are “blocked” which is curated and maintained by the Ministry [to this day, no such resource is publicly available however, AIW has a list of online resources that are regularly measured for blocking with [OONI Probe](https://ooni.org/install/) monitored relying on OONI for blocking]. Independent legal experts [believe](https://rm.coe.int/azerbaijan-analysis-of-legislation-on-access-to-information-december-2/16808ae03c) this kind of authority is restrictive in nature, particularly since it forces Internet Service Providers (ISPs) in the country to block access to specific online resources. According to the law, the Ministry of Transport, Hich Technologies and Communication is the executive authority deciding on the type of information that is relevant, which websites get blocked and what information must be removed and so on. Yet again illustrating the state’s control and involvement as the sole decision making body. ### Internet censorship and media freedom environment -In Azerbaijan, the media is viewed as an element in the state’s political struggle according to the journalist Shahin Hajiyev. In an interview with Voice of America, Hajiyev [said](https://www.voanews.com/press-freedom/space-free-media-azerbaijan-shrinking-journalists-say), “The space for free media is shrinking by each passing year.”  +In Azerbaijan, the media is viewed as an element in the state’s political struggle according to the journalist Shahin Hajiyev. In an interview with Voice of America, Hajiyev [said](https://www.voanews.com/press-freedom/space-free-media-azerbaijan-shrinking-journalists-say), “The space for free media is shrinking by each passing year.” -As a result, almost all of the independent media platforms are currently based outside Azerbaijan with teams dispatched and operating on the ground, delivering news for an audience inside and outside Azerbaijan. Most of their websites have been blocked in the country since 2017. At the time, a court in Baku [ruled](https://freedomhouse.org/article/azerbaijani-court-bans-independent-media-websites) to block access to several popular independent and opposition websites including Meydan TV, Radio Liberty’s Azerbaijani service and three others. According to most recent measurements as presented in this report, the original list now includes at least ten news websites. Scores of independent journalists and activists have [reported](https://www.codastory.com/authoritarian-tech/surveillance-and-internet-disruption-in-baku/) over the years of [hacking attempts](https://www.az-netwatch.org/news/hacking-alert-activists-and-journalists-targeted-online-ongoing/) on their social media accounts and profiles, for their outspokenness or activism. Just as many have [faced criminal proceedings](https://www.az-netwatch.org/news/6-journalists-and-bloggers-behind-bars-in-azerbaijan/) in real life. As a result of deliberate internet censorship and stifling of online dissent, Azerbaijan consistently ranked poorly on a number of international rightswatch organisations, including the annual Freedom House, Freedom on the Net [report](https://freedomhouse.org/country/azerbaijan/freedom-net/2020). +As a result, almost all of the independent media platforms are currently based outside Azerbaijan with teams dispatched and operating on the ground, delivering news for an audience inside and outside Azerbaijan. Most of their websites have been blocked in the country since 2017. At the time, a court in Baku [ruled](https://freedomhouse.org/article/azerbaijani-court-bans-independent-media-websites) to block access to several popular independent and opposition websites including Meydan TV, Radio Liberty’s Azerbaijani service and three others. According to most recent measurements as presented in this report, the original list now includes at least ten news websites. Scores of independent journalists and activists have [reported](https://www.codastory.com/authoritarian-tech/surveillance-and-internet-disruption-in-baku/) over the years of [hacking attempts](https://www.az-netwatch.org/news/hacking-alert-activists-and-journalists-targeted-online-ongoing/) on their social media accounts and profiles, for their outspokenness or activism. Just as many have [faced criminal proceedings](https://www.az-netwatch.org/news/6-journalists-and-bloggers-behind-bars-in-azerbaijan/) in real life. As a result of deliberate internet censorship and stifling of online dissent, Azerbaijan consistently ranked poorly on a number of international rightswatch organisations, including the annual Freedom House, Freedom on the Net [report](https://freedomhouse.org/country/azerbaijan/freedom-net/2020). -In addition to controlling key infrastructure, the powers vested in the government, the lack of freedoms and the [country’s dismal human rights record](https://www.amnesty.org/en/countries/europe-and-central-asia/azerbaijan/), have also made the government a key player in controlling the internet and the content shared online. The government also resorted to a vast array of [information controls](https://advox.globalvoices.org/2018/01/03/will-we-ever-see-the-end-of-information-control-in-azerbaijan/) over the years, in great part thanks to [authoritarian technology](https://www.opendemocracy.net/en/odr/azerbaijans-authoritarianism-goes-digital/) it has [purchased](https://www.bloomberg.com/news/articles/2020-10-08/sandvine-s-tools-used-for-web-censoring-in-more-than-a-dozen-nations?sref=zAhiqZgK) over [the years](https://www.occrp.org/en/daily/4136-azerbaijan-bought-hacking-team-s-surveillance-spyware-leaks-reveal) and the legal acts such as the 2016 Law on Licenses and Permits and the Resolution No. 174 on "additional conditions required for the issuance of special permits (licenses) depending on the nature of the activity", granting government institutions such as the State Security Services powers to snoop around its citizens both online and offline. +In addition to controlling key infrastructure, the powers vested in the government, the lack of freedoms and the [country’s dismal human rights record](https://www.amnesty.org/en/countries/europe-and-central-asia/azerbaijan/), have also made the government a key player in controlling the internet and the content shared online. The government also resorted to a vast array of [information controls](https://advox.globalvoices.org/2018/01/03/will-we-ever-see-the-end-of-information-control-in-azerbaijan/) over the years, in great part thanks to [authoritarian technology](https://www.opendemocracy.net/en/odr/azerbaijans-authoritarianism-goes-digital/) it has [purchased](https://www.bloomberg.com/news/articles/2020-10-08/sandvine-s-tools-used-for-web-censoring-in-more-than-a-dozen-nations?sref=zAhiqZgK) over [the years](https://www.occrp.org/en/daily/4136-azerbaijan-bought-hacking-team-s-surveillance-spyware-leaks-reveal) and the legal acts such as the 2016 Law on Licenses and Permits and the Resolution No. 174 on "additional conditions required for the issuance of special permits (licenses) depending on the nature of the activity", granting government institutions such as the State Security Services powers to snoop around its citizens both online and offline. -Before advanced censorship authoritarian technology was [available](https://eurasianet.org/azerbaijan-state-and-dissidents-acquire-new-weapons-cyber-war), the government used its arbitrary powers to restrict the internet. This practice began as early as 2003, when the government requested ISPs to [cut access](https://www.refworld.org/docid/46e69182c.html) to all Chechen websites. The same year, now former, the Ministry of Communications and High Technologies official [said](https://www.refworld.org/docid/46e69182c.html), the government body had the right to ask ISPs to block sites containing material that was ‘contrary to the mentality, traditions, and customs’ of the country. +Before advanced censorship authoritarian technology was [available](https://eurasianet.org/azerbaijan-state-and-dissidents-acquire-new-weapons-cyber-war), the government used its arbitrary powers to restrict the internet. This practice began as early as 2003, when the government requested ISPs to [cut access](https://www.refworld.org/docid/46e69182c.html) to all Chechen websites. The same year, now former, the Ministry of Communications and High Technologies official [said](https://www.refworld.org/docid/46e69182c.html), the government body had the right to ask ISPs to block sites containing material that was ‘contrary to the mentality, traditions, and customs’ of the country. -Since then, [blocking access to websites](https://www.az-netwatch.org/news/the-tale-of-blocked-websites/) has become a norm, with the government relying on the existing law on “Information, Informatisation and access to Information”, to implement internet carry out online censorship. That, in addition, non-transparency of the existing process in which ISPs and mobile operators implement government orders to block access to specific content online, makes it hard to fight internet censorship and advocate for internet freedoms. In the absence of any legal remedies at home, often users are left with little or no solution mechanisms.   +Since then, [blocking access to websites](https://www.az-netwatch.org/news/the-tale-of-blocked-websites/) has become a norm, with the government relying on the existing law on “Information, Informatisation and access to Information”, to implement internet carry out online censorship. That, in addition, non-transparency of the existing process in which ISPs and mobile operators implement government orders to block access to specific content online, makes it hard to fight internet censorship and advocate for internet freedoms. In the absence of any legal remedies at home, often users are left with little or no solution mechanisms. ## Findings -As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) (between 1st January 2020 to 1st May 2021), we found a number of independent news media websites and circumvention tool websites blocked throughout the analysis period. We also found social media websites and apps temporarily blocked amid the 2020 Nagorno-Karabakh war. We share further details in the sections below. +As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ) (between 1st January 2020 to 1st May 2021), we found a number of independent news media websites and circumvention tool websites blocked throughout the analysis period. We also found social media websites and apps temporarily blocked amid the 2020 Nagorno-Karabakh war. We share further details in the sections below. ### Blocked news media websites @@ -199,8 +199,8 @@ These domains include: 1. `azerbaycansaati.tv` 2. `criminal.az` -3. `www.24saat.org`  -4. `www.abzas.net`  +3. `www.24saat.org` +4. `www.abzas.net` 5. `www.azadliq.info` 6. `www.azadliq.org` 7. `www.gununsesi.info` @@ -208,111 +208,111 @@ These domains include: 9. `www.kanal13.tv` 10. `www.meydan.tv` -Independent Azerbaijani media websites (such as `azadliq.info` and `meydan.tv`) have [reportedly](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/) been blocked since early 2017 for “posing a threat” to Azerbaijan’s national security. The state prosecutor reportedly accused these websites of sharing content that promotes violence, hatred, extremism, violates privacy and constitutes slander. However, the blocking of these media outlets may have been [politically motivated](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/). Azadliq, for example, reported on the [business dealings](https://www.azadliq.org/a/intihar-sui-qesd/28097510.html) of Vice President Mehriban Aliyeva's private foundation, while the website of Gunun Sesi (which has [reportedly](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) been blocked since August 2018) is [operated](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) by Parviz Hashimli, a former political prisoner.   +Independent Azerbaijani media websites (such as `azadliq.info` and `meydan.tv`) have [reportedly](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/) been blocked since early 2017 for “posing a threat” to Azerbaijan’s national security. The state prosecutor reportedly accused these websites of sharing content that promotes violence, hatred, extremism, violates privacy and constitutes slander. However, the blocking of these media outlets may have been [politically motivated](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/). Azadliq, for example, reported on the [business dealings](https://www.azadliq.org/a/intihar-sui-qesd/28097510.html) of Vice President Mehriban Aliyeva's private foundation, while the website of Gunun Sesi (which has [reportedly](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) been blocked since August 2018) is [operated](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) by Parviz Hashimli, a former political prisoner. -OONI data also suggests that the site (`www.occrp.org`) of the [Organized Crime and Corruption Reporting Project (OCCRP)](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&domain=www.occrp.org&probe_cc=AZ) and the site (`www.rferl.org`) of [RadioFreeEurope/RadioLiberty (RFE/RFL)](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&domain=www.rferl.org&probe_cc=AZ&only=anomalies) were blocked in Azerbaijan as well. The blocking of the OCCRP site [reportedly](https://www.meydan.tv/en/article/occrp-blocked-in-azerbaijan/) began in September 2017, following the publication of a major investigation (“[Azerbaijani Laundromat](https://www.occrp.org/en/azerbaijanilaundromat/)”) into corruption, bribery, and money laundering in which powerful figures were allegedly involved. The blocking of the RFE/RFL website also [reportedly](https://www.rferl.org/a/azerbaijan-rferl-service-website-court-orders-blocked/28482679.html) began in 2017, following an Azerbaijani court order which RFE/RFL [described](https://www.rferl.org/a/azerbaijan-rferl-service-website-court-orders-blocked/28482679.html) as “another blatant attempt at silencing its reporting in the country”.   +OONI data also suggests that the site (`www.occrp.org`) of the [Organized Crime and Corruption Reporting Project (OCCRP)](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&domain=www.occrp.org&probe_cc=AZ) and the site (`www.rferl.org`) of [RadioFreeEurope/RadioLiberty (RFE/RFL)](https://explorer.ooni.org/search?until=2021-06-03&since=2021-05-04&domain=www.rferl.org&probe_cc=AZ&only=anomalies) were blocked in Azerbaijan as well. The blocking of the OCCRP site [reportedly](https://www.meydan.tv/en/article/occrp-blocked-in-azerbaijan/) began in September 2017, following the publication of a major investigation (“[Azerbaijani Laundromat](https://www.occrp.org/en/azerbaijanilaundromat/)”) into corruption, bribery, and money laundering in which powerful figures were allegedly involved. The blocking of the RFE/RFL website also [reportedly](https://www.rferl.org/a/azerbaijan-rferl-service-website-court-orders-blocked/28482679.html) began in 2017, following an Azerbaijani court order which RFE/RFL [described](https://www.rferl.org/a/azerbaijan-rferl-service-website-court-orders-blocked/28482679.html) as “another blatant attempt at silencing its reporting in the country”. The following chart illustrates OONI measurement findings from the testing of these media websites (on multiple local networks) in Azerbaijan between January 2020 to May 2021. ![](images/image18.png) -Chart: Blocking of media websites in Azerbaijan between January 2020 to May 2021 based on OONI data, [https://explorer.ooni.org/search?since=2020-01-01&probe\_cc=AZ&test\_name=web\_connectivity&only=anomalies](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies)  +**Chart:** Blocking of media websites in Azerbaijan between January 2020 to May 2021 based on OONI data, [https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies) -Most OONI measurements collected from the testing of these media websites (on multiple local networks) [presented anomalies](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&probe_cc=AZ&test_name=web_connectivity&only=anomalies) throughout the testing period, strongly suggesting potential blocking. In contrast, most of the other URLs tested from the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists) during this period were found [accessible](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&probe_cc=AZ&test_name=web_connectivity) (on the same networks in Azerbaijan). +Most OONI measurements collected from the testing of these media websites (on multiple local networks) [presented anomalies](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&probe_cc=AZ&test_name=web_connectivity&only=anomalies) throughout the testing period, strongly suggesting potential blocking. In contrast, most of the other URLs tested from the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists) during this period were found [accessible](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&probe_cc=AZ&test_name=web_connectivity) (on the same networks in Azerbaijan). What is evident from the above chart is that most of the anomalous measurements (for these media websites) consistently presented the same “HTTP failures” throughout the testing period. This means that OONI Probe was able to successfully resolve via DNS the IP addresses of these domains and successfully establish TCP connections to the resolved IP addresses, but when it attempted to retrieve the website content, the connection failed. In most cases, we observe that these media websites appear to be blocked because the HTTP requests time out. This is an indication that the blocking might be implemented by means of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi). -The above chart aggregates measurement findings (per domain) from many different networks in Azerbaijan (as annotated in the chart, where the largest dots correspond to measurements collected from 9 local networks). The reason why we see different measurement results (such as “OK” and “HTTP failure”) for the same URL on the same dates is because those results were collected from different networks. This therefore suggests that media website blocking varies from network to network in Azerbaijan. This is further evident through a per [ASN](https://ooni.org/support/glossary/#asn) analysis of measurements. +The above chart aggregates measurement findings (per domain) from many different networks in Azerbaijan (as annotated in the chart, where the largest dots correspond to measurements collected from 9 local networks). The reason why we see different measurement results (such as “OK” and “HTTP failure”) for the same URL on the same dates is because those results were collected from different networks. This therefore suggests that media website blocking varies from network to network in Azerbaijan. This is further evident through a per [ASN](https://ooni.org/support/glossary/#asn) analysis of measurements. For example, when looking at measurements collected from the testing of `azerbaycansaati.tv`, we can see that it presented signs of blocking on at least 6 local networks (`AS15723`, `AS28787`, `AS31721`, `AS34170`, `AS39232`, `AS42779`) in February 2020, but had seemingly been unblocked on `AS15723` by May 2020 (as all subsequent measurements showed that the site was accessible on those networks), as illustrated through the following chart. ![](images/image9.png) -Chart: Blocking of `azerbaycansaati.tv` across AS networks in Azerbaijan between February 2020 to April 2021 based on OONI data, [https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=azerbaycansaati.tv&probe\_cc=AZ](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=azerbaycansaati.tv&probe_cc=AZ)  +**Chart:** Blocking of `azerbaycansaati.tv` across AS networks in Azerbaijan between February 2020 to April 2021 based on OONI data, [https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=azerbaycansaati.tv&probe_cc=AZ](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=azerbaycansaati.tv&probe_cc=AZ) When looking at measurements collected from the testing of `www.azadliq.info`, we can see that it was found blocked on many networks, while accessible on a few networks (such as `AS29584` and `AS15723`), as illustrated below. ![](images/image12.png) -Chart: Blocking of `azadliq.info` across AS networks in Azerbaijan between February 2020 to April 2021 based on OONI data, [https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.azadliq.info&probe\_cc=AZ](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.azadliq.info&probe_cc=AZ)  +**Chart:** Blocking of `azadliq.info` across AS networks in Azerbaijan between February 2020 to April 2021 based on OONI data, [https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.azadliq.info&probe\_cc=AZ](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.azadliq.info&probe_cc=AZ) Interestingly, `www.gununsesi.org` appears to remain blocked on most networks, but `www.gununsesi.info` was found accessible every time it was tested (on multiple networks) from 26th July 2020 onwards (even though it previously presented signs of blocking, similarly to `www.gununsesi.org`), as illustrated through the following chart. ![](images/image10.png) -Chart: Blocking of `gununsesi.info` across AS networks in Azerbaijan between February 2020 to April 2021 based on OONI data, [https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.gununsesi.info&probe\_cc=AZ](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.gununsesi.info&probe_cc=AZ)  +**Chart:** Blocking of `gununsesi.info` across AS networks in Azerbaijan between February 2020 to April 2021 based on OONI data, [https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.gununsesi.info&probe\_cc=AZ](https://explorer.ooni.org/search?until=2021-06-04&since=2021-05-05&domain=www.gununsesi.info&probe_cc=AZ) -When `www.gununsesi.org` is accessed from a browser, it redirects to `www.gununsesi.info`, suggesting that most internet users in Azerbaijan likely do not experience the blocking of `www.gununsesi.org` (at least on the networks where `www.gununsesi.info` is accessible). +When `www.gununsesi.org` is accessed from a browser, it redirects to `www.gununsesi.info`, suggesting that most internet users in Azerbaijan likely do not experience the blocking of `www.gununsesi.org` (at least on the networks where `www.gununsesi.info` is accessible). ### Blocking of social media amid 2020 Nagorno-Karabakh war ![](images/image7.png) -War [erupted](https://www.bbc.com/news/world-europe-54324772) in [Nagorno-Karabakh](https://www.bbc.com/news/world-europe-18270325) on 27th September 2020 between Azerbaijan and ethnic Armenians, once Azerbaijan [launched an offensive](https://www.theguardian.com/world/2020/nov/10/nagorno-karabakh-armenia-pm-signs-deal-to-end-war-with-azerbaijan-and-russia) to retake Nagorno-Karabakh and surrounding regions under Armenian control. The war lasted for six weeks -- up until 10th November 2020, when a Russian-brokered truce was signed. +War [erupted](https://www.bbc.com/news/world-europe-54324772) in [Nagorno-Karabakh](https://www.bbc.com/news/world-europe-18270325) on 27th September 2020 between Azerbaijan and ethnic Armenians, once Azerbaijan [launched an offensive](https://www.theguardian.com/world/2020/nov/10/nagorno-karabakh-armenia-pm-signs-deal-to-end-war-with-azerbaijan-and-russia) to retake Nagorno-Karabakh and surrounding regions under Armenian control. The war lasted for six weeks -- up until 10th November 2020, when a Russian-brokered truce was signed. -The peace deal is [viewed](https://www.bbc.com/news/world-europe-54882564) as a victory for Azerbaijan (which regained most of the occupied territories surrounding Nagorno-Karabakh and large parts of Nagorno-Karabakh), and a defeat for Armenia. +The peace deal is [viewed](https://www.bbc.com/news/world-europe-54882564) as a victory for Azerbaijan (which regained most of the occupied territories surrounding Nagorno-Karabakh and large parts of Nagorno-Karabakh), and a defeat for Armenia. -This was the second war in the disputed region, following the [first Nagorno-Karabakh war](https://www.rferl.org/a/1065626.html) between 1988 to 1994, which ended with a ceasefire and tens of thousands of deaths. The region has since been internationally recognized as Azerbaijani, but run by separatist ethnic Armenians (backed by the Armenian government). While the second conflict ended with a peace deal and fewer casualties in comparison to the first war, the new ceasefire agreement prompted anger and [protests](https://www.crisisgroup.org/europe-central-asia/caucasus/nagorno-karabakh-conflict/getting-ceasefire-peace-nagorno-karabakh) in Armenia, while the details of the new peace deal have been [described](https://www.theguardian.com/world/2020/nov/10/nagorno-karabakh-armenia-pm-signs-deal-to-end-war-with-azerbaijan-and-russia) by the International Crisis Group as rather vague. There are therefore [concerns](https://www.cfr.org/global-conflict-tracker/conflict/nagorno-karabakh-conflict) of cease-fire violations and renewed tensions without successful mediation efforts. +This was the second war in the disputed region, following the [first Nagorno-Karabakh war](https://www.rferl.org/a/1065626.html) between 1988 to 1994, which ended with a ceasefire and tens of thousands of deaths. The region has since been internationally recognized as Azerbaijani, but run by separatist ethnic Armenians (backed by the Armenian government). While the second conflict ended with a peace deal and fewer casualties in comparison to the first war, the new ceasefire agreement prompted anger and [protests](https://www.crisisgroup.org/europe-central-asia/caucasus/nagorno-karabakh-conflict/getting-ceasefire-peace-nagorno-karabakh) in Armenia, while the details of the new peace deal have been [described](https://www.theguardian.com/world/2020/nov/10/nagorno-karabakh-armenia-pm-signs-deal-to-end-war-with-azerbaijan-and-russia) by the International Crisis Group as rather vague. There are therefore [concerns](https://www.cfr.org/global-conflict-tracker/conflict/nagorno-karabakh-conflict) of cease-fire violations and renewed tensions without successful mediation efforts. #### Blocking of social media websites -Amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), OONI data [shows](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies&until=2021-06-10) that access to several social media websites and apps was blocked in Azerbaijan. The following chart, limited to social media websites that presented signs of blocking between February 2020 to May 2021, aggregates OONI measurement findings collected from 4 AS networks in Azerbaijan. +Amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), OONI data [shows](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies&until=2021-06-10) that access to several social media websites and apps was blocked in Azerbaijan. The following chart, limited to social media websites that presented signs of blocking between February 2020 to May 2021, aggregates OONI measurement findings collected from 4 AS networks in Azerbaijan. ![](images/image5.png) -Chart: Blocking of social media websites in Azerbaijan based on OONI data (collected between March 2020 to May 2021), [https://explorer.ooni.org/search?since=2020-01-01&probe\_cc=AZ&test\_name=web\_connectivity&only=anomalies](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies)  +**Chart:** Blocking of social media websites in Azerbaijan based on OONI data (collected between March 2020 to May 2021), [https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies) -As is evident from the above chart, most of these social media websites primarily [presented signs of blocking](https://explorer.ooni.org/measurement/20201030T232015Z_webconnectivity_AZ_34876_n1_UWIBPqC9nIZq1L28?input=http%3A%2F%2Fwww.facebook.com%2F) during the 2020 Nagorno-Karabakh war (between 27th September 2020 to 10th November 2020), but were found [accessible](https://explorer.ooni.org/measurement/20201224T085701Z_webconnectivity_AZ_196925_n1_gILh6TXHAwl86poo?input=http%3A%2F%2Fwww.facebook.com%2F) when tested (on several networks in Azerbaijan) in the months before and after the war. Notably, most anomalous measurements presented HTTP failures (because the HTTP requests timed out), similarly to the blocking of news media websites (discussed previously). This provides a stronger indication that these social media websites were blocked, particularly since ISPs often use the same censorship technique(s) to block a variety of different websites. +As is evident from the above chart, most of these social media websites primarily [presented signs of blocking](https://explorer.ooni.org/measurement/20201030T232015Z_webconnectivity_AZ_34876_n1_UWIBPqC9nIZq1L28?input=http%3A%2F%2Fwww.facebook.com%2F) during the 2020 Nagorno-Karabakh war (between 27th September 2020 to 10th November 2020), but were found [accessible](https://explorer.ooni.org/measurement/20201224T085701Z_webconnectivity_AZ_196925_n1_gILh6TXHAwl86poo?input=http%3A%2F%2Fwww.facebook.com%2F) when tested (on several networks in Azerbaijan) in the months before and after the war. Notably, most anomalous measurements presented HTTP failures (because the HTTP requests timed out), similarly to the blocking of news media websites (discussed previously). This provides a stronger indication that these social media websites were blocked, particularly since ISPs often use the same censorship technique(s) to block a variety of different websites. -Two social media websites though (`www.viber.com` and `vimeo.com`) appear to have potentially been blocked in dates before and after the 2020 Nagorno-Karabakh war. OONI data shows that the [testing](https://explorer.ooni.org/search?since=2021-05-11&domain=www.viber.com&probe_cc=AZ) of `www.viber.com` presented signs of blocking on 3 AS networks (`AS28787`, `AS50274`, `AS57923`) from as early as February and March 2020, as illustrated through the following chart. +Two social media websites though (`www.viber.com` and `vimeo.com`) appear to have potentially been blocked in dates before and after the 2020 Nagorno-Karabakh war. OONI data shows that the [testing](https://explorer.ooni.org/search?since=2021-05-11&domain=www.viber.com&probe_cc=AZ) of `www.viber.com` presented signs of blocking on 3 AS networks (`AS28787`, `AS50274`, `AS57923`) from as early as February and March 2020, as illustrated through the following chart. ![](images/image1.png) -Chart: Testing of `www.viber.com` across AS networks in Azerbaijan between February 2020 to May 2021 based on OONI data, [https://explorer.ooni.org/search?since=2021-05-11&domain=www.viber.com&probe\_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&domain=www.viber.com&probe_cc=AZ)  +**Chart:** Testing of `www.viber.com` across AS networks in Azerbaijan between February 2020 to May 2021 based on OONI data, [https://explorer.ooni.org/search?since=2021-05-11&domain=www.viber.com&probe\_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&domain=www.viber.com&probe_cc=AZ) More recent measurements collected from `AS28787` suggest that `www.viber.com` is accessible, but as the site hadn’t been tested on `AS50274` and `AS57923` since March 2020, it remains unclear if the potential block was persistent on those networks. Amid the 2020 Nagorno-Karabakh war, we observe the potential blocking of `www.viber.com` on the `AS60258` and `AS34876` networks. However, the limited testing coverage throughout this period limits our ability to evaluate whether the site was blocked on other networks, when the blocking started, and how long it lasted. Testing coverage of `www.viber.com` increased in January 2021, enabling us to confirm the accessibility of the site on multiple AS networks (as demonstrated through the above chart). -Vimeo.com, on the other hand, only started presenting signs of potential blocking after the 2020 Nagorno-Karabakh war. The following chart shows that vimeo.com appeared to have been [blocked](https://explorer.ooni.org/search?since=2020-11-01&domain=vimeo.com&probe_cc=AZ) in Azerbaijan between 1st December 2020 to 11th January 2021, though the blocking may have started at an earlier date (we cannot exclude this hypothesis due to the limited testing coverage). +Vimeo.com, on the other hand, only started presenting signs of potential blocking after the 2020 Nagorno-Karabakh war. The following chart shows that vimeo.com appeared to have been [blocked](https://explorer.ooni.org/search?since=2020-11-01&domain=vimeo.com&probe_cc=AZ) in Azerbaijan between 1st December 2020 to 11th January 2021, though the blocking may have started at an earlier date (we cannot exclude this hypothesis due to the limited testing coverage). ![](images/image16.png) -Chart: Blocking of vimeo.com in Azerbaijan between December 2020 to January 2021 based on OONI data. +**Chart:** Blocking of vimeo.com in Azerbaijan between December 2020 to January 2021 based on OONI data. -This is suggested from OONI measurements which [show](https://explorer.ooni.org/measurement/20210111T075445Z_webconnectivity_AZ_39232_n1_MwN4fOSghJ1ttluE?input=https%3A%2F%2Fvimeo.com%2F) that connections to `https://vimeo.com` timed out when performing the HTTPS request to the site [every time that it was tested](https://explorer.ooni.org/search?since=2020-11-01&domain=vimeo.com&probe_cc=AZ) throughout this period. This is further suggested by the fact that we observe the exact same HTTP failures and timeouts (on 5 AS networks) as we did for the testing of other websites (discussed previously). The following chart provides a per-ASN breakdown of `vimeo.com` measurement coverage in Azerbaijan between February 2020 to May 2021. +This is suggested from OONI measurements which [show](https://explorer.ooni.org/measurement/20210111T075445Z_webconnectivity_AZ_39232_n1_MwN4fOSghJ1ttluE?input=https%3A%2F%2Fvimeo.com%2F) that connections to `https://vimeo.com` timed out when performing the HTTPS request to the site [every time that it was tested](https://explorer.ooni.org/search?since=2020-11-01&domain=vimeo.com&probe_cc=AZ) throughout this period. This is further suggested by the fact that we observe the exact same HTTP failures and timeouts (on 5 AS networks) as we did for the testing of other websites (discussed previously). The following chart provides a per-ASN breakdown of `vimeo.com` measurement coverage in Azerbaijan between February 2020 to May 2021. ![](images/image15.png) -Chart: Testing of `vimeo.com` across AS networks in Azerbaijan between February 2020 to May 2021 based on OONI data, [https://explorer.ooni.org/search?since=2021-05-11&domain=vimeo.com&probe\_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&domain=vimeo.com&probe_cc=AZ)  +**Chart:** Testing of `vimeo.com` across AS networks in Azerbaijan between February 2020 to May 2021 based on OONI data, [https://explorer.ooni.org/search?since=2021-05-11&domain=vimeo.com&probe_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&domain=vimeo.com&probe_cc=AZ) -Currently, vimeo.com appears to be [accessible](https://explorer.ooni.org/search?since=2020-11-01&domain=vimeo.com&probe_cc=AZ) on several networks in Azerbaijan, as illustrated through the above chart. +Currently, vimeo.com appears to be [accessible](https://explorer.ooni.org/search?since=2020-11-01&domain=vimeo.com&probe_cc=AZ) on several networks in Azerbaijan, as illustrated through the above chart. #### Blocking of instant messaging apps -We also observe the potential blocking of instant messaging apps in Azerbaijan during the 2020 Nagorno-Karabakh war. The following chart aggregates OONI measurements collected from the testing of [WhatsApp](https://ooni.org/nettest/whatsapp/), [Telegram](https://ooni.org/nettest/telegram/), and [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/) (on up to 9 AS networks) in Azerbaijan between January 2020 to May 2021. +We also observe the potential blocking of instant messaging apps in Azerbaijan during the 2020 Nagorno-Karabakh war. The following chart aggregates OONI measurements collected from the testing of [WhatsApp](https://ooni.org/nettest/whatsapp/), [Telegram](https://ooni.org/nettest/telegram/), and [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/) (on up to 9 AS networks) in Azerbaijan between January 2020 to May 2021. ![](images/image17.png) -Chart: Blocking of instant messaging apps in Azerbaijan based on OONI data (collected between January 2020 to May 2021), [https://explorer.ooni.org/search?since=2021-05-11&probe\_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&probe_cc=AZ)  +**Chart:** Blocking of instant messaging apps in Azerbaijan based on OONI data (collected between January 2020 to May 2021), [https://explorer.ooni.org/search?since=2021-05-11&probe_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&probe_cc=AZ) -It is evident from the above chart that the testing of [WhatsApp](https://explorer.ooni.org/measurement/20201013T091449Z_AS57293_FzfUoxzPn0MJ54QB0MfBDrvkise9gIC9s6UiCR9zBR3Xid4vPk) and [Telegram](https://explorer.ooni.org/measurement/20201013T091504Z_AS57293_EQwZAGeu31VSogUgZGcSD7dWkTXeO1GazIa0qYIEe3X8Sui5mB) primarily presented anomalies between September 2020 to November 2020, which coincides with the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). During this period, most measurements presented anomalies and signs of [WhatsApp](https://explorer.ooni.org/measurement/20201022T224244Z_whatsapp_AZ_57293_n1_URgnFAYE4EVt5qIP) and [Telegram](https://explorer.ooni.org/measurement/20201022T223851Z_telegram_AZ_57293_n1_k6k8ioloAMiEnxi5) blocking. But most measurements collected before and after the war show that these apps were [reachable](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=whatsapp) from local networks in Azerbaijan. While we also observe some [anomalous measurements pertaining to the testing of Facebook Messenger](https://explorer.ooni.org/measurement/20201029T142707Z_facebookmessenger_AZ_28787_n1_AVXqhgPois0hQj9z) during the 2020 Nagorno-Karabakh war, they are much fewer in comparison to the anomalies presented in the testing of Telegram and WhatsApp, limiting our confidence with respect to Facebook Messenger blocking. +It is evident from the above chart that the testing of [WhatsApp](https://explorer.ooni.org/measurement/20201013T091449Z_AS57293_FzfUoxzPn0MJ54QB0MfBDrvkise9gIC9s6UiCR9zBR3Xid4vPk) and [Telegram](https://explorer.ooni.org/measurement/20201013T091504Z_AS57293_EQwZAGeu31VSogUgZGcSD7dWkTXeO1GazIa0qYIEe3X8Sui5mB) primarily presented anomalies between September 2020 to November 2020, which coincides with the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). During this period, most measurements presented anomalies and signs of [WhatsApp](https://explorer.ooni.org/measurement/20201022T224244Z_whatsapp_AZ_57293_n1_URgnFAYE4EVt5qIP) and [Telegram](https://explorer.ooni.org/measurement/20201022T223851Z_telegram_AZ_57293_n1_k6k8ioloAMiEnxi5) blocking. But most measurements collected before and after the war show that these apps were [reachable](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=whatsapp) from local networks in Azerbaijan. While we also observe some [anomalous measurements pertaining to the testing of Facebook Messenger](https://explorer.ooni.org/measurement/20201029T142707Z_facebookmessenger_AZ_28787_n1_AVXqhgPois0hQj9z) during the 2020 Nagorno-Karabakh war, they are much fewer in comparison to the anomalies presented in the testing of Telegram and WhatsApp, limiting our confidence with respect to Facebook Messenger blocking. -It is possible that other social media apps may have been blocked during the 2020 Nagorno-Karabakh war, but our findings are limited to the testing of [WhatsApp](https://ooni.org/nettest/whatsapp/), [Telegram](https://ooni.org/nettest/telegram/), and [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/) because these are the primary social media apps that [OONI Probe](https://ooni.org/install/) provides tests for. [Signal measurements](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=signal) are excluded from this analysis, because the [OONI Probe Signal test](https://ooni.org/nettest/signal/) was only released recently. However, recent measurements collected from Azerbaijan suggest that the Signal Private Messenger app [works](https://explorer.ooni.org/measurement/20210611T105112Z_signal_AZ_31721_n1_ZnEFy0jHnlAzE4X3) on several networks. +It is possible that other social media apps may have been blocked during the 2020 Nagorno-Karabakh war, but our findings are limited to the testing of [WhatsApp](https://ooni.org/nettest/whatsapp/), [Telegram](https://ooni.org/nettest/telegram/), and [Facebook Messenger](https://ooni.org/nettest/facebook-messenger/) because these are the primary social media apps that [OONI Probe](https://ooni.org/install/) provides tests for. [Signal measurements](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=signal) are excluded from this analysis, because the [OONI Probe Signal test](https://ooni.org/nettest/signal/) was only released recently. However, recent measurements collected from Azerbaijan suggest that the Signal Private Messenger app [works](https://explorer.ooni.org/measurement/20210611T105112Z_signal_AZ_31721_n1_ZnEFy0jHnlAzE4X3) on several networks. ##### WhatsApp -Between 27th September 2020 to 11th November 2020, OONI measurements consistently presented signs of WhatsApp blocking in Azerbaijan, as illustrated through [OONI Probe WhatsApp measurements](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp) collected from several local networks. During this period, all [OONI Probe WhatsApp tests showed signs of WhatsApp blocking](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp&only=anomalies), while all [other](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp) OONI measurements collected before and after this time period successfully connected to the tested endpoints of WhatsApp’s mobile app and web interface. +Between 27th September 2020 to 11th November 2020, OONI measurements consistently presented signs of WhatsApp blocking in Azerbaijan, as illustrated through [OONI Probe WhatsApp measurements](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp) collected from several local networks. During this period, all [OONI Probe WhatsApp tests showed signs of WhatsApp blocking](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp&only=anomalies), while all [other](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp) OONI measurements collected before and after this time period successfully connected to the tested endpoints of WhatsApp’s mobile app and web interface. -Within the [anomalous measurements](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp&only=anomalies) (which suggested WhatsApp blocking), we observe that attempted connections to WhatsApp's registration service and web interface (web.whatsapp.com) failed. It’s worth noting that in [some cases](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT), the HTTP request to web.whatsapp.com succeeded, while the HTTPS request failed, as illustrated below. +Within the [anomalous measurements](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=whatsapp&only=anomalies) (which suggested WhatsApp blocking), we observe that attempted connections to WhatsApp's registration service and web interface (web.whatsapp.com) failed. It’s worth noting that in [some cases](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT), the HTTP request to web.whatsapp.com succeeded, while the HTTPS request failed, as illustrated below. ![](images/image14.png) -This could be an indication of SNI based filtering of WhatsApp. We also observe that the TLS handshakes field presented failures, further suggesting that access to WhatsApp was blocked in Azerbaijan by means of SNI based filtering. This pattern was observed on several ISPs in Azerbaijan between 27th September 2020 to 11th November 2020. +This could be an indication of SNI based filtering of WhatsApp. We also observe that the TLS handshakes field presented failures, further suggesting that access to WhatsApp was blocked in Azerbaijan by means of SNI based filtering. This pattern was observed on several ISPs in Azerbaijan between 27th September 2020 to 11th November 2020. ##### Telegram -Similarly to the testing of WhatsApp, OONI measurements presented [signs of Telegram blocking](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=telegram&only=anomalies) in Azerbaijan between 27th September 2020 to 11th November 2020. But unlike WhatsApp, a few Telegram tests during this period were successful. This is suggested through [OONI Probe Telegram measurements](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=telegram) collected from several local networks in Azerbaijan (illustrated through the previous chart) between 1st September 2020 to 27th November 2020. +Similarly to the testing of WhatsApp, OONI measurements presented [signs of Telegram blocking](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=telegram&only=anomalies) in Azerbaijan between 27th September 2020 to 11th November 2020. But unlike WhatsApp, a few Telegram tests during this period were successful. This is suggested through [OONI Probe Telegram measurements](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=telegram) collected from several local networks in Azerbaijan (illustrated through the previous chart) between 1st September 2020 to 27th November 2020. -We observe similarities between the date range of potential blocking (of WhatsApp and Telegram), but we see that both HTTP and HTTPS requests to Telegram Web (web.telegram.org) [timed out](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp). We observe a [timeout in the TLS handshake](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=telegram&only=anomalies), suggesting TLS level interference of Telegram. +We observe similarities between the date range of potential blocking (of WhatsApp and Telegram), but we see that both HTTP and HTTPS requests to Telegram Web (web.telegram.org) [timed out](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp). We observe a [timeout in the TLS handshake](https://explorer.ooni.org/search?until=2020-11-27&since=2020-09-01&probe_cc=AZ&test_name=telegram&only=anomalies), suggesting TLS level interference of Telegram. As a result, it is possible that internet users in Azerbaijan couldn’t use the Telegram and WhatsApp mobile apps (on the tested networks) between 27th September 2020 to 11th November 2020. @@ -322,74 +322,74 @@ Numerous circumvention tool websites presented signs of potential blocking when ![](images/image3.png) -Chart: Blocking of circumvention tool websites in Azerbaijan based on OONI data (collected between January 2020 to May 2021), [https://explorer.ooni.org/search?since=2020-01-01&probe\_cc=AZ&test\_name=web\_connectivity&only=anomalies](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies)  +**Chart:** Blocking of circumvention tool websites in Azerbaijan based on OONI data (collected between January 2020 to May 2021), [https://explorer.ooni.org/search?since=2020-01-01&probe\_cc=AZ&test\_name=web\_connectivity&only=anomalies](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies) -Similarly to the blocking of news media and social media websites, we observe that the testing of circumvention tool websites [often](https://explorer.ooni.org/measurement/20210120T135659Z_webconnectivity_AZ_196961_n1_mmv1ohvJMITxyfXX?input=https%3A%2F%2Fwww.torproject.org%2F) [resulted](https://explorer.ooni.org/measurement/20201229T213913Z_webconnectivity_AZ_196925_n1_GrvcacpX2MhB9VN1?input=https%3A%2F%2Fpsiphon.ca%2F) in [HTTP failures](https://explorer.ooni.org/measurement/20210111T075445Z_webconnectivity_AZ_39232_n1_MwN4fOSghJ1ttluE?input=https%3A%2F%2Fwww.tunnelbear.com%2F) caused by connection timeouts. This consistency in terms of failures, observed on several AS networks over the period of a year, strongly suggests blocking of these circumvention tool websites. As testing coverage increased from January 2021 onwards, we observed an increased volume of anomalous measurements, most presenting the same HTTP failures. +Similarly to the blocking of news media and social media websites, we observe that the testing of circumvention tool websites [often](https://explorer.ooni.org/measurement/20210120T135659Z_webconnectivity_AZ_196961_n1_mmv1ohvJMITxyfXX?input=https%3A%2F%2Fwww.torproject.org%2F) [resulted](https://explorer.ooni.org/measurement/20201229T213913Z_webconnectivity_AZ_196925_n1_GrvcacpX2MhB9VN1?input=https%3A%2F%2Fpsiphon.ca%2F) in [HTTP failures](https://explorer.ooni.org/measurement/20210111T075445Z_webconnectivity_AZ_39232_n1_MwN4fOSghJ1ttluE?input=https%3A%2F%2Fwww.tunnelbear.com%2F) caused by connection timeouts. This consistency in terms of failures, observed on several AS networks over the period of a year, strongly suggests blocking of these circumvention tool websites. As testing coverage increased from January 2021 onwards, we observed an increased volume of anomalous measurements, most presenting the same HTTP failures. It’s worth highlighting though that these circumvention tool sites did not display signs of blocking on all tested networks, suggesting that their blocking varies from ISP to ISP in Azerbaijan (as appears to also be the case for other blocked sites, discussed previously). Nor did they display signs of blocking every time that they were tested on those networks over time. All of this is evident from the above chart that aggregates measurement results (per URL) from several different AS networks in Azerbaijan, displaying a variance (“OK” and “anomalous”) in measurement results. -Even though we observe consistency in terms of blocking techniques adopted by ISPs, we see that different sites are blocked by different ISPs at different moments in time. Some circumvention tool sites (such as [torproject.org](https://explorer.ooni.org/measurement/20210527T135516Z_webconnectivity_AZ_39232_n1_OoeoxKSNHIoW7Qhd?input=https%3A%2F%2Fwww.torproject.org%2F) and [purevpn.com](https://explorer.ooni.org/measurement/20210526T172253Z_webconnectivity_AZ_39232_n1_mbREiLo5qZDnnnKG?input=https%3A%2F%2Fwww.purevpn.com%2F)) were recently found accessible on some networks. On the other hand, sites like [nordvpn.com](https://explorer.ooni.org/measurement/20210524T135523Z_webconnectivity_AZ_39232_n1_Nm8am7CasoIu9veI?input=https%3A%2F%2Fnordvpn.com%2F), [protonvpn.com](https://explorer.ooni.org/measurement/20210527T135516Z_webconnectivity_AZ_39232_n1_OoeoxKSNHIoW7Qhd?input=https%3A%2F%2Fprotonvpn.com%2F), [psiphon.ca](https://explorer.ooni.org/measurement/20210531T134250Z_webconnectivity_AZ_39232_n1_I7J0JPg8Hjr1ENeq?input=https%3A%2F%2Fpsiphon.ca%2F), and [strongvpn.com](https://explorer.ooni.org/measurement/20210527T124850Z_webconnectivity_AZ_39232_n1_wu1VodRzK1Jsloki?input=https%3A%2F%2Fstrongvpn.com%2F) presented HTTP failures (caused by connection timeouts) in most measurements collected in April 2021 -- providing a stronger signal of potential blocking. +Even though we observe consistency in terms of blocking techniques adopted by ISPs, we see that different sites are blocked by different ISPs at different moments in time. Some circumvention tool sites (such as [torproject.org](https://explorer.ooni.org/measurement/20210527T135516Z_webconnectivity_AZ_39232_n1_OoeoxKSNHIoW7Qhd?input=https%3A%2F%2Fwww.torproject.org%2F) and [purevpn.com](https://explorer.ooni.org/measurement/20210526T172253Z_webconnectivity_AZ_39232_n1_mbREiLo5qZDnnnKG?input=https%3A%2F%2Fwww.purevpn.com%2F)) were recently found accessible on some networks. On the other hand, sites like [nordvpn.com](https://explorer.ooni.org/measurement/20210524T135523Z_webconnectivity_AZ_39232_n1_Nm8am7CasoIu9veI?input=https%3A%2F%2Fnordvpn.com%2F), [protonvpn.com](https://explorer.ooni.org/measurement/20210527T135516Z_webconnectivity_AZ_39232_n1_OoeoxKSNHIoW7Qhd?input=https%3A%2F%2Fprotonvpn.com%2F), [psiphon.ca](https://explorer.ooni.org/measurement/20210531T134250Z_webconnectivity_AZ_39232_n1_I7J0JPg8Hjr1ENeq?input=https%3A%2F%2Fpsiphon.ca%2F), and [strongvpn.com](https://explorer.ooni.org/measurement/20210527T124850Z_webconnectivity_AZ_39232_n1_wu1VodRzK1Jsloki?input=https%3A%2F%2Fstrongvpn.com%2F) presented HTTP failures (caused by connection timeouts) in most measurements collected in April 2021 -- providing a stronger signal of potential blocking. Given that a number of news media websites appear to have been blocked in Azerbaijan throughout the testing period (between January 2020 to May 2021), it is possible that ISPs in Azerbaijan may have blocked access to popular circumvention tool sites in an attempt to prevent internet users from circumventing internet censorship (such as the blocking of news media websites). #### Tor and Psiphon -As OONI Probe only includes tests for 3 circumvention tools ([Tor](https://ooni.org/nettest/tor/), [Psiphon](https://ooni.org/nettest/psiphon/), [RiseupVPN](https://ooni.org/nettest/riseupvpn/)), it remains unclear if the apps of all the other aforementioned circumvention tool websites were blocked as well (i.e. it is possible that only their websites were blocked). [RiseupVPN measurements](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=riseupvpn) are excluded from this analysis (since this [test](https://ooni.org/nettest/riseupvpn/) was only released a few months ago), but recent measurements collected from Azerbaijan suggest that RiseupVPN [works](https://explorer.ooni.org/measurement/20210611T095059Z_riseupvpn_AZ_196925_n1_dJNOgZV4x6UvCb41) on local networks. +As OONI Probe only includes tests for 3 circumvention tools ([Tor](https://ooni.org/nettest/tor/), [Psiphon](https://ooni.org/nettest/psiphon/), [RiseupVPN](https://ooni.org/nettest/riseupvpn/)), it remains unclear if the apps of all the other aforementioned circumvention tool websites were blocked as well (i.e. it is possible that only their websites were blocked). [RiseupVPN measurements](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=riseupvpn) are excluded from this analysis (since this [test](https://ooni.org/nettest/riseupvpn/) was only released a few months ago), but recent measurements collected from Azerbaijan suggest that RiseupVPN [works](https://explorer.ooni.org/measurement/20210611T095059Z_riseupvpn_AZ_196925_n1_dJNOgZV4x6UvCb41) on local networks. -The following chart aggregates all OONI measurements collected from the [testing of Psiphon](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=psiphon) and [Tor](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor) on up to 9 AS networks in Azerbaijan, between January 2020 to May 2021. +The following chart aggregates all OONI measurements collected from the [testing of Psiphon](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=psiphon) and [Tor](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor) on up to 9 AS networks in Azerbaijan, between January 2020 to May 2021. ![](images/image13.png) -Chart: Testing of circumvention tools in Azerbaijan based on OONI data (collected between January 2020 to May 2021), [https://explorer.ooni.org/search?since=2021-05-11&probe\_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&probe_cc=AZ)  +**Chart:** Testing of circumvention tools in Azerbaijan based on OONI data (collected between January 2020 to May 2021), [https://explorer.ooni.org/search?since=2021-05-11&probe\_cc=AZ](https://explorer.ooni.org/search?since=2021-05-11&probe_cc=AZ) -As is evident, the [testing of Tor](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor) presented a significantly larger volume of anomalies in comparison to the [testing of Psiphon](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=psiphon) throughout the testing period. We also see that both Tor and Psiphon presented a larger volume of anomalous measurements between September 2020 to November 2020, which coincides with the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). As access to social media platforms was [blocked](https://www.balcanicaucaso.org/eng/Areas/Azerbaijan/Azerbaijan-the-Internet-in-times-of-war-205919) in Azerbaijan during this period (as discussed previously), it’s possible that internet users may have made [greater use of circumvention tools](https://surfshark.com/blog/azerbaijan-starts-blocking-media) (to bypass social media blocks), which local ISPs may have attempted to block. +As is evident, the [testing of Tor](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor) presented a significantly larger volume of anomalies in comparison to the [testing of Psiphon](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=psiphon) throughout the testing period. We also see that both Tor and Psiphon presented a larger volume of anomalous measurements between September 2020 to November 2020, which coincides with the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). As access to social media platforms was [blocked](https://www.balcanicaucaso.org/eng/Areas/Azerbaijan/Azerbaijan-the-Internet-in-times-of-war-205919) in Azerbaijan during this period (as discussed previously), it’s possible that internet users may have made [greater use of circumvention tools](https://surfshark.com/blog/azerbaijan-starts-blocking-media) (to bypass social media blocks), which local ISPs may have attempted to block. - In [Tor anomalous measurements](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor&only=anomalies), we observe a blocking attempt by several local ISPs, but internet users in Azerbaijan could likely use [Tor](https://www.torproject.org/) nonetheless. Using Tor requires connecting to 1 Tor directory authority, and the [OONI Probe Tor test](https://ooni.org/nettest/tor/) measures the reachability of 10 Tor directory authorities, in addition to 15 Tor Browser [bridges](https://bridges.torproject.org/) (which help with circumventing Tor blocking). +In [Tor anomalous measurements](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor&only=anomalies), we observe a blocking attempt by several local ISPs, but internet users in Azerbaijan could likely use [Tor](https://www.torproject.org/) nonetheless. Using Tor requires connecting to 1 Tor directory authority, and the [OONI Probe Tor test](https://ooni.org/nettest/tor/) measures the reachability of 10 Tor directory authorities, in addition to 15 Tor Browser [bridges](https://bridges.torproject.org/) (which help with circumventing Tor blocking). -In OONI measurements, we see that some ISPs in Azerbaijan appear to have blocked connections to a subset of Tor directory authorities (for example, on `AS28787` we see that [connections failed to 7](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL) out of 10 directory authorities, while on `AS42779` we see that [connections failed to 8](https://explorer.ooni.org/measurement/20201014T142943Z_AS42779_dEF2KNECjwt7r8LqeuxgPNnRDJrnDZ6nss3kQnu47NXWydCDWk) out of 10 directory authorities). On some networks (such as `AS196925`), we see [all connections to all 10 tested directory authorities failing](https://explorer.ooni.org/measurement/20201029T211052Z_tor_AZ_196925_n1_ZgoVCny0Au4972rz) (as well as failed connections to 8 Tor bridges), while on other networks (such as `AS39397`), we see [successful connections to all 10 tested directory authorities](https://explorer.ooni.org/measurement/20201030T205554Z_tor_AZ_39397_n1_GEwUsMJXNrLnnDcG) (where there’s no sign of Tor blocking). +In OONI measurements, we see that some ISPs in Azerbaijan appear to have blocked connections to a subset of Tor directory authorities (for example, on `AS28787` we see that [connections failed to 7](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL) out of 10 directory authorities, while on `AS42779` we see that [connections failed to 8](https://explorer.ooni.org/measurement/20201014T142943Z_AS42779_dEF2KNECjwt7r8LqeuxgPNnRDJrnDZ6nss3kQnu47NXWydCDWk) out of 10 directory authorities). On some networks (such as `AS196925`), we see [all connections to all 10 tested directory authorities failing](https://explorer.ooni.org/measurement/20201029T211052Z_tor_AZ_196925_n1_ZgoVCny0Au4972rz) (as well as failed connections to 8 Tor bridges), while on other networks (such as `AS39397`), we see [successful connections to all 10 tested directory authorities](https://explorer.ooni.org/measurement/20201030T205554Z_tor_AZ_39397_n1_GEwUsMJXNrLnnDcG) (where there’s no sign of Tor blocking). -In short, it seems that some ISPs in Azerbaijan attempted to block access to Tor (some more effectively than others), while other ISPs did not. Even in the cases where ISPs appear to have attempted to block access to Tor, the fact that they often didn’t block access to all Tor directory authorities suggests that internet users in Azerbaijan may have been able to use Tor nonetheless (since connecting to 1 Tor directory authority is sufficient for using Tor). Moreover, it’s possible to circumvent Tor blocking through the use of [Tor bridges](https://bridges.torproject.org/) (many of which were [not blocked](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL), according to OONI measurements). Tor also maintains a list of fallback directory authorities (which are not currently tested by the [OONI Probe Tor test](https://ooni.org/nettest/tor/)) and a connection to any of those is usually sufficient to successfully bootstrap Tor. +In short, it seems that some ISPs in Azerbaijan attempted to block access to Tor (some more effectively than others), while other ISPs did not. Even in the cases where ISPs appear to have attempted to block access to Tor, the fact that they often didn’t block access to all Tor directory authorities suggests that internet users in Azerbaijan may have been able to use Tor nonetheless (since connecting to 1 Tor directory authority is sufficient for using Tor). Moreover, it’s possible to circumvent Tor blocking through the use of [Tor bridges](https://bridges.torproject.org/) (many of which were [not blocked](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL), according to OONI measurements). Tor also maintains a list of fallback directory authorities (which are not currently tested by the [OONI Probe Tor test](https://ooni.org/nettest/tor/)) and a connection to any of those is usually sufficient to successfully bootstrap Tor. -The hypothesis that the blocking attempt against Tor was ineffective is further suggested by Tor Metrics, which [show a large spike in directly connecting Tor users from Azerbaijan](https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off) during this period (amid the 2020 Nagorno-Karabakh war). +The hypothesis that the blocking attempt against Tor was ineffective is further suggested by Tor Metrics, which [show a large spike in directly connecting Tor users from Azerbaijan](https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off) during this period (amid the 2020 Nagorno-Karabakh war). ![](images/image20.png) -Chart: Tor Metrics, Directly connecting users from Azerbaijan between 1st August 2020 to 30th December 2020, [https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off](https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off)  +**Chart:** Tor Metrics, Directly connecting users from Azerbaijan between 1st August 2020 to 30th December 2020, [https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off](https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off) -Even in cases where ISPs appeared to [attempt to block access to Tor](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL), on most networks they did not block access to all Tor directory authorities, meaning that internet users in Azerbaijan could likely use Tor anyway -- as further suggested by the spike in Tor usage during this time frame. +Even in cases where ISPs appeared to [attempt to block access to Tor](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL), on most networks they did not block access to all Tor directory authorities, meaning that internet users in Azerbaijan could likely use Tor anyway -- as further suggested by the spike in Tor usage during this time frame. -Quite similarly, Tor Metrics also [show a large spike in Tor bridge users from Azerbaijan](https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-12-30&country=az) during the same period. +Quite similarly, Tor Metrics also [show a large spike in Tor bridge users from Azerbaijan](https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-12-30&country=az) during the same period. ![](images/image2.png) -Chart: Tor Metrics, Bridge users from Azerbaijan between 1st August 2020 to 30th December 2020, [https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-12-30&country=az](https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-12-30&country=az)  +**Chart:** Tor Metrics, Bridge users from Azerbaijan between 1st August 2020 to 30th December 2020, [https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-12-30&country=az](https://metrics.torproject.org/userstats-bridge-country.html?start=2020-08-01&end=2020-12-30&country=az) -As we observe the [potential blocking of all 10 tested Tor directory authorities](https://explorer.ooni.org/measurement/20201029T211052Z_tor_AZ_196925_n1_ZgoVCny0Au4972rz) on some networks, this may have resulted in the increased use of Tor bridges in Azerbaijan. Recent OONI measurements suggest that [attempts to block Tor appear to be ongoing](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor), while Tor’s website (www.torproject.org) appears to be [blocked](https://explorer.ooni.org/measurement/20210530T105053Z_webconnectivity_AZ_60258_n1_H3bQBQuydpgs0Gjm?input=https%3A%2F%2Fwww.torproject.org%2F) on some networks, but [accessible](https://explorer.ooni.org/measurement/20210611T095506Z_webconnectivity_AZ_196925_n1_4oaAaCs3Qxn2z8ZB?input=https%3A%2F%2Fwww.torproject.org%2F) on many others. +As we observe the [potential blocking of all 10 tested Tor directory authorities](https://explorer.ooni.org/measurement/20201029T211052Z_tor_AZ_196925_n1_ZgoVCny0Au4972rz) on some networks, this may have resulted in the increased use of Tor bridges in Azerbaijan. Recent OONI measurements suggest that [attempts to block Tor appear to be ongoing](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=tor), while Tor’s website (www.torproject.org) appears to be [blocked](https://explorer.ooni.org/measurement/20210530T105053Z_webconnectivity_AZ_60258_n1_H3bQBQuydpgs0Gjm?input=https%3A%2F%2Fwww.torproject.org%2F) on some networks, but [accessible](https://explorer.ooni.org/measurement/20210611T095506Z_webconnectivity_AZ_196925_n1_4oaAaCs3Qxn2z8ZB?input=https%3A%2F%2Fwww.torproject.org%2F) on many others. -Psiphon also presented an increased volume of [anomalous measurements](https://explorer.ooni.org/measurement/20201014T041710Z_AS203622_dzuDmZBynGUPa8JvJS4Tq6xYe4VWpIHr3BAm3yQQfetESvbGRD) amid the 2020 Nagorno-Karabakh, suggesting that local ISPs may have attempted to block access to it as well. However, we also observe many [successful measurements](https://explorer.ooni.org/measurement/20201013T092707Z_AS57293_psovJca5TFn72zSu5eXtXF9bF4KAUFoYXR8O8FVRm271RF9wXB) during this period (where the [OONI Probe Psiphon test](https://ooni.org/nettest/psiphon/) is able to successfully bootstrap Psiphon and fetch a webpage from the internet), and [Psiphon has built-in circumvention techniques (fallback options)](https://github.com/Psiphon-Labs/psiphon-tunnel-core#circumvention-optimizations), suggesting that attempts to block Psiphon may have been quite ineffective.   +Psiphon also presented an increased volume of [anomalous measurements](https://explorer.ooni.org/measurement/20201014T041710Z_AS203622_dzuDmZBynGUPa8JvJS4Tq6xYe4VWpIHr3BAm3yQQfetESvbGRD) amid the 2020 Nagorno-Karabakh, suggesting that local ISPs may have attempted to block access to it as well. However, we also observe many [successful measurements](https://explorer.ooni.org/measurement/20201013T092707Z_AS57293_psovJca5TFn72zSu5eXtXF9bF4KAUFoYXR8O8FVRm271RF9wXB) during this period (where the [OONI Probe Psiphon test](https://ooni.org/nettest/psiphon/) is able to successfully bootstrap Psiphon and fetch a webpage from the internet), and [Psiphon has built-in circumvention techniques (fallback options)](https://github.com/Psiphon-Labs/psiphon-tunnel-core#circumvention-optimizations), suggesting that attempts to block Psiphon may have been quite ineffective. -While the testing of Psiphon presented a few anomalies after the 2020 Nagorno-Karabakh war, those anomalies are likely [false positives](https://ooni.org/support/faq/#what-are-false-positives) (particularly since the vast majority of measurements were successful). Recent measurements suggest that access to Psiphon’s website (`psiphon.ca`) is [blocked](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=web_connectivity&domain=psiphon.ca) on several networks, but the Psiphon app seems to [work](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=psiphon) (based on tests on several networks). +While the testing of Psiphon presented a few anomalies after the 2020 Nagorno-Karabakh war, those anomalies are likely [false positives](https://ooni.org/support/faq/#what-are-false-positives) (particularly since the vast majority of measurements were successful). Recent measurements suggest that access to Psiphon’s website (`psiphon.ca`) is [blocked](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=web_connectivity&domain=psiphon.ca) on several networks, but the Psiphon app seems to [work](https://explorer.ooni.org/search?until=2021-06-12&since=2021-05-13&probe_cc=AZ&test_name=psiphon)(based on tests on several networks). ## Conclusion -Press freedom appears to be quite limited in Azerbaijan, as suggested by the [blocking of several independent news media websites](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies&until=2021-06-15) in the country. These media websites presented signs of blocking throughout their testing (on several local AS networks) between January 2020 to May 2021 (corroborating [past reports](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/) on the blocking of media websites in Azerbaijan), with recent OONI measurements suggesting that their [blocking remains ongoing](https://explorer.ooni.org/measurement/20210611T125048Z_webconnectivity_AZ_196925_n1_UPy97oSW7tyjVHUP?input=https%3A%2F%2Fwww.gununsesi.org%2F). +Press freedom appears to be quite limited in Azerbaijan, as suggested by the [blocking of several independent news media websites](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies&until=2021-06-15) in the country. These media websites presented signs of blocking throughout their testing (on several local AS networks) between January 2020 to May 2021 (corroborating [past reports](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/) on the blocking of media websites in Azerbaijan), with recent OONI measurements suggesting that their [blocking remains ongoing](https://explorer.ooni.org/measurement/20210611T125048Z_webconnectivity_AZ_196925_n1_UPy97oSW7tyjVHUP?input=https%3A%2F%2Fwww.gununsesi.org%2F). -Potentially in an attempt to prevent the circumvention of media censorship, ISPs in Azerbaijan appear to have [blocked access to a number of circumvention tool websites](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies&until=2021-06-15) over the last year as well. It remains unclear, however, if the apps of these circumvention tool sites were also blocked (as they were not tested as part of this study); and even if they were, it’s possible that local internet users may have been able to use them nonetheless, given that circumvention tools often include built-in circumvention techniques for evading censors. +Potentially in an attempt to prevent the circumvention of media censorship, ISPs in Azerbaijan appear to have [blocked access to a number of circumvention tool websites](https://explorer.ooni.org/search?since=2020-01-01&probe_cc=AZ&test_name=web_connectivity&only=anomalies&until=2021-06-15) over the last year as well. It remains unclear, however, if the apps of these circumvention tool sites were also blocked (as they were not tested as part of this study); and even if they were, it’s possible that local internet users may have been able to use them nonetheless, given that circumvention tools often include built-in circumvention techniques for evading censors. -Amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), ISPs in Azerbaijan appear to have attempted to block the [Tor](https://explorer.ooni.org/measurement/20201029T211052Z_tor_AZ_196925_n1_ZgoVCny0Au4972rz) and [Psiphon](https://explorer.ooni.org/measurement/20201014T041710Z_AS203622_dzuDmZBynGUPa8JvJS4Tq6xYe4VWpIHr3BAm3yQQfetESvbGRD) circumvention tools. Yet, these attempts were likely ineffective, given that both tools have [built-in circumvention techniques](https://github.com/Psiphon-Labs/psiphon-tunnel-core#circumvention-optimizations) and [fallback options](https://gitlab.torproject.org/legacy/trac/-/wikis/doc/FallbackDirectoryMirrors) for circumventing blocks. In [Tor measurements](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL), we observe that most ISPs did not block all tested Tor directory authorities, suggesting that it was possible to use [Tor](https://www.torproject.org/) nonetheless (as also indicated by the [spike in Tor usage](https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off) from Azerbaijan during that period). Similarly, many Psiphon measurements during this period were [successful](https://explorer.ooni.org/measurement/20201013T092707Z_AS57293_psovJca5TFn72zSu5eXtXF9bF4KAUFoYXR8O8FVRm271RF9wXB), suggesting that it may have been possible to use the [Psiphon VPN](https://psiphon.ca/) on many networks. +Amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), ISPs in Azerbaijan appear to have attempted to block the [Tor](https://explorer.ooni.org/measurement/20201029T211052Z_tor_AZ_196925_n1_ZgoVCny0Au4972rz) and [Psiphon](https://explorer.ooni.org/measurement/20201014T041710Z_AS203622_dzuDmZBynGUPa8JvJS4Tq6xYe4VWpIHr3BAm3yQQfetESvbGRD) circumvention tools. Yet, these attempts were likely ineffective, given that both tools have [built-in circumvention techniques](https://github.com/Psiphon-Labs/psiphon-tunnel-core#circumvention-optimizations) and [fallback options](https://gitlab.torproject.org/legacy/trac/-/wikis/doc/FallbackDirectoryMirrors) for circumventing blocks. In [Tor measurements](https://explorer.ooni.org/measurement/20201014T142140Z_AS28787_C6HQmTGrA24EbrxJSHhmaEQ47D4QJZqAlTPWR5FNTJbRp97eXL), we observe that most ISPs did not block all tested Tor directory authorities, suggesting that it was possible to use [Tor](https://www.torproject.org/) nonetheless (as also indicated by the [spike in Tor usage](https://metrics.torproject.org/userstats-relay-country.html?start=2020-08-01&end=2020-12-30&country=az&events=off) from Azerbaijan during that period). Similarly, many Psiphon measurements during this period were [successful](https://explorer.ooni.org/measurement/20201013T092707Z_AS57293_psovJca5TFn72zSu5eXtXF9bF4KAUFoYXR8O8FVRm271RF9wXB), suggesting that it may have been possible to use the [Psiphon VPN](https://psiphon.ca/) on many networks. -Several social media websites (such as `www.facebook.com` and `www.youtube.com`) and apps (primarily [WhatsApp](https://explorer.ooni.org/measurement/20201022T224244Z_whatsapp_AZ_57293_n1_URgnFAYE4EVt5qIP) and [Telegram](https://explorer.ooni.org/measurement/20201022T223851Z_telegram_AZ_57293_n1_k6k8ioloAMiEnxi5)) presented [signs of blocking](https://explorer.ooni.org/measurement/20201030T232015Z_webconnectivity_AZ_34876_n1_UWIBPqC9nIZq1L28?input=http%3A%2F%2Fwww.facebook.com%2F) between September 2020 to November 2020, which coincides with the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). It is therefore possible that ISPs may have increased efforts to block circumvention tools (during this period) in an attempt to prevent the circumvention of social media censorship. +Several social media websites (such as `www.facebook.com` and `www.youtube.com`) and apps (primarily [WhatsApp](https://explorer.ooni.org/measurement/20201022T224244Z_whatsapp_AZ_57293_n1_URgnFAYE4EVt5qIP) and [Telegram](https://explorer.ooni.org/measurement/20201022T223851Z_telegram_AZ_57293_n1_k6k8ioloAMiEnxi5)) presented [signs of blocking](https://explorer.ooni.org/measurement/20201030T232015Z_webconnectivity_AZ_34876_n1_UWIBPqC9nIZq1L28?input=http%3A%2F%2Fwww.facebook.com%2F) between September 2020 to November 2020, which coincides with the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772). It is therefore possible that ISPs may have increased efforts to block circumvention tools (during this period) in an attempt to prevent the circumvention of social media censorship. Interestingly, we observe similar censorship techniques adopted by different ISPs in Azerbaijan, but variance in terms of which internet services are blocked by ISPs over time. In other words, we see ISPs blocking websites and apps in similar ways (seemingly using the same censorship techniques), but different ISPs block access to different websites and apps (and sometimes this varies at different moments in time). -Throughout the testing period, independent news media and circumvention tool websites presented HTTP failures caused by connection timeouts, suggesting the potential use of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi) by ISPs in Azerbaijan. Similarly, when social media websites were temporarily blocked amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), their testing also presented HTTP failures caused by connection timeouts. This suggests that most ISPs in Azerbaijan block websites using similar (if not the same) censorship techniques. +Throughout the testing period, independent news media and circumvention tool websites presented HTTP failures caused by connection timeouts, suggesting the potential use of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi) by ISPs in Azerbaijan. Similarly, when social media websites were temporarily blocked amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), their testing also presented HTTP failures caused by connection timeouts. This suggests that most ISPs in Azerbaijan block websites using similar (if not the same) censorship techniques. -Both WhatsApp and Telegram presented signs of TLS level interference on several different AS networks in Azerbaijan amid the 2020 Nagorno-Karabakh war. In the case of [WhatsApp](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT), the HTTP requests to `web.whatsapp.com` succeeded, while the HTTPS requests failed (during the TLS handshake), which could be an indication of SNI based filtering. In the case of [Telegram](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp), we see that both HTTP and HTTPS requests to `web.telegram.org` timed out.  +Both WhatsApp and Telegram presented signs of TLS level interference on several different AS networks in Azerbaijan amid the 2020 Nagorno-Karabakh war. In the case of [WhatsApp](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT), the HTTP requests to `web.whatsapp.com` succeeded, while the HTTPS requests failed (during the TLS handshake), which could be an indication of SNI based filtering. In the case of [Telegram](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp), we see that both HTTP and HTTPS requests to `web.telegram.org` timed out. As media censorship (and the blocking of circumvention tool websites) appears to be ongoing in Azerbaijan, there is need for further testing to evaluate these censorship events in more depth over time. The temporary blocking of social media amid the 2020 Nagorno-Karabakh war also suggests that new censorship events can emerge in Azerbaijan, as political events evolve. -This study was carried out through the use of [open methodologies](https://github.com/ooni/spec/tree/master/nettests), [free and open source software](https://ooni.org/install/), and [open data](https://ooni.org/data/), enabling independent third-party verification of our research findings. We encourage researchers to expand upon this study by [running OONI Probe](https://ooni.org/install/) and analyzing [OONI measurements from Azerbaijan](https://explorer.ooni.org/search?since=2021-05-16&probe_cc=AZ).     +This study was carried out through the use of [open methodologies](https://github.com/ooni/spec/tree/master/nettests), [free and open source software](https://ooni.org/install/), and [open data](https://ooni.org/data/), enabling independent third-party verification of our research findings. We encourage researchers to expand upon this study by [running OONI Probe](https://ooni.org/install/) and analyzing [OONI measurements from Azerbaijan](https://explorer.ooni.org/search?since=2021-05-16&probe_cc=AZ). ## Acknowledgements -We thank all [OONI Probe](https://ooni.org/install/) users in Azerbaijan for contributing measurements, making this study possible. +We thank all [OONI Probe](https://ooni.org/install/) users in Azerbaijan for contributing measurements, making this study possible. diff --git a/content/post/2021-how-signal-private-messenger-blocked-around-the-world/index.md b/content/post/2021-how-signal-private-messenger-blocked-around-the-world/index.md index 999a31d83..8a983e6a2 100644 --- a/content/post/2021-how-signal-private-messenger-blocked-around-the-world/index.md +++ b/content/post/2021-how-signal-private-messenger-blocked-around-the-world/index.md @@ -3,7 +3,7 @@ title: "How countries attempt to block Signal Private Messenger App around the w description: "This report shares OONI measurements on the blocking of the Signal Private Messenger app in Iran, China, Cuba, and Uzbekistan." author: "Maria Xynou, Arturo Filastò" date: "2021-10-21" -tags: ["iran", "china", "cuba", "uzbekistan", "censorship", "signal", "country-ir", "country-cn", "country-cu", "country-uz"] +tags: ["iran", "china", "cuba", "uzbekistan", "censorship", "signal", "country-ir", "country-cn", "country-cu", "country-uz", "theme-im"] categories: ["report"] --- diff --git a/content/post/2021-multiperspective-view-internet-censorship-myanmar/index.md b/content/post/2021-multiperspective-view-internet-censorship-myanmar/index.md index e19aacf01..e031c3745 100644 --- a/content/post/2021-multiperspective-view-internet-censorship-myanmar/index.md +++ b/content/post/2021-multiperspective-view-internet-censorship-myanmar/index.md @@ -3,7 +3,7 @@ title: "A multi-perspective view of Internet censorship in Myanmar" description: "This research paper examines internet censorship in Myanmar following the 2021 military coup." author: "Maria Xynou" date: "2021-11-11" -tags: ["myanmar", "censorship", "blackout", "military-coup", "country-mm", "paper"] +tags: ["myanmar", "censorship", "blackout", "military-coup", "country-mm", "paper", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2021-myanmar-internet-blocks-and-outages/index.md b/content/post/2021-myanmar-internet-blocks-and-outages/index.md index b7c730fec..ab4f7fe7f 100644 --- a/content/post/2021-myanmar-internet-blocks-and-outages/index.md +++ b/content/post/2021-myanmar-internet-blocks-and-outages/index.md @@ -3,7 +3,7 @@ title: "Myanmar: Data on internet blocks and internet outages following military description: "This report shares OONI measurements on internet blocks in Myanmar, and IODA data on the internet outages following the military coup." author: "Maria Xynou (OONI), Ramakrishna Padmanabhan (CAIDA, UC San Diego), Phyu Kyaw, Myanmar ICT for Development Organization (MIDO), Arturo Filastò (OONI)" date: "2021-03-09" -tags: ["myanmar", "censorship", "blackout", "military-coup", "country-mm"] +tags: ["myanmar", "censorship", "blackout", "military-coup", "country-mm", "theme-news_media", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2021-no-access-lgbtiq-website-censorship-six-countries/index.md b/content/post/2021-no-access-lgbtiq-website-censorship-six-countries/index.md index 8b0016e19..f9aea3632 100644 --- a/content/post/2021-no-access-lgbtiq-website-censorship-six-countries/index.md +++ b/content/post/2021-no-access-lgbtiq-website-censorship-six-countries/index.md @@ -3,7 +3,7 @@ title: "No Access: LGBTIQ Website Censorship in Six Countries" description: "This research report shares OONI measurements on the blocking of LGBTIQ websites in Malaysia, Indonesia, Russia, Iran, Saudi Arabia, and the UAE, along with qualitative findings from interviews with LGBTIQ communities." author: "Open Observatory of Network Interference (OONI), OutRight Action International, The Citizen Lab" date: "2021-08-31" -tags: ["lgbtiq", "censorship", "country-my", "country-id", "country-ru", "country-ir", "country-sa", "country-ae"] +tags: ["lgbtiq", "censorship", "country-my", "country-id", "country-ru", "country-ir", "country-sa", "country-ae", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/2021-russia-blocks-tor/index.md b/content/post/2021-russia-blocks-tor/index.md index fdaebb531..123c941a9 100644 --- a/content/post/2021-russia-blocks-tor/index.md +++ b/content/post/2021-russia-blocks-tor/index.md @@ -3,7 +3,7 @@ title: "Russia started blocking Tor" description: "This report shares OONI measurements on the blocking of Tor and the Tor Project's website in Russia." author: "Maria Xynou, Arturo Filastò" date: "2021-12-17" -tags: ["russia", "censorship", "tor", "country-ru"] +tags: ["russia", "censorship", "tor", "country-ru", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2021-uganda-general-election-blocks-and-outage/index.md b/content/post/2021-uganda-general-election-blocks-and-outage/index.md index 1c8b2a374..c72826008 100644 --- a/content/post/2021-uganda-general-election-blocks-and-outage/index.md +++ b/content/post/2021-uganda-general-election-blocks-and-outage/index.md @@ -3,7 +3,7 @@ title: "Uganda: Data on internet blocks and nationwide internet outage amid 2021 description: "This report shares OONI measurements on the blocking of social media in Uganda, and IODA data on the 4-day internet outage amid its 2021 general election." author: "Maria Xynou (OONI), Simone Basso (OONI), Ramakrishna Padmanabhan (CAIDA, UC San Diego), Arturo Filastò (OONI), DefendDefenders, Defenders Protection Initiative" date: "2021-01-22" -tags: ["uganda", "censorship", "blackout", "elections", "country-ug"] +tags: ["uganda", "censorship", "blackout", "elections", "country-ug", "theme-im", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2021-zambia-social-media-blocks-amid-elections/index.md b/content/post/2021-zambia-social-media-blocks-amid-elections/index.md index cbcc1245c..6e0c970c3 100644 --- a/content/post/2021-zambia-social-media-blocks-amid-elections/index.md +++ b/content/post/2021-zambia-social-media-blocks-amid-elections/index.md @@ -3,7 +3,7 @@ title: "Zambia: Social media blocked amid 2021 general elections" description: "This report shares OONI measurements on the blocking of social media in Zambia amid its 2021 general election." author: "Maria Xynou, Arturo Filastò" date: "2021-08-24" -tags: ["zambia", "censorship", "elections", "country-zm"] +tags: ["zambia", "censorship", "elections", "country-zm", "theme-im", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/2022-azerbaijan-and-armenia-blocks-tiktok/index.md b/content/post/2022-azerbaijan-and-armenia-blocks-tiktok/index.md index 955962774..3a8ab29de 100644 --- a/content/post/2022-azerbaijan-and-armenia-blocks-tiktok/index.md +++ b/content/post/2022-azerbaijan-and-armenia-blocks-tiktok/index.md @@ -3,13 +3,13 @@ title: "Azerbaijan and Armenia block TikTok amid border clashes" description: "This report shares OONI data on the blocking of TikTok in Azerbaijan and Armenia amid border clashes in September 2022." author: "Arturo Filastò (OONI), Arzu Geybulla (Azerbaijan Internet Watch), Maria Xynou (OONI)" date: "2022-09-16" -tags: ["azerbaijan", "censorship", "country-az", "country-am"] +tags: ["azerbaijan", "censorship", "country-az", "country-am", "theme-social_media"] categories: ["report"] --- -Earlier this week, on 12th September 2022, fighting [erupted](https://www.bbc.com/news/world-europe-62828239) between Azerbaijani and Armenian troops along their border. Over the next few days, community members in Azerbaijan reported that the TikTok app was blocked locally. +Earlier this week, on 12th September 2022, fighting [erupted](https://www.bbc.com/news/world-europe-62828239) between Azerbaijani and Armenian troops along their border. Over the next few days, community members in Azerbaijan reported that the TikTok app was blocked locally. -We analyzed [OONI network measurement data](https://ooni.org/data) to investigate the block. We found that TikTok has been blocked in both [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM&test_name=web_connectivity&domain=tiktok.com) over the last few days. +We analyzed [OONI network measurement data](https://ooni.org/data) to investigate the block. We found that TikTok has been blocked in both [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM&test_name=web_connectivity&domain=tiktok.com) over the last few days. In this report, we share our technical findings. In both Armenia and Azerbaijan, we found TLS and DNS level interference of TikTok domains and endpoints during the border clashes. @@ -17,23 +17,23 @@ In this report, we share our technical findings. In both Armenia and Azerbaijan, ## Background -On 12th September 2022, fighting [erupted](https://www.bbc.com/news/world-europe-62888891) along the Azerbaijan-Armenia border between Azerbaijani and Armenian troops. The next day, Russia brokered a [ceasefire](https://www.bbc.com/news/world-europe-62828239), but it was immediately breached. Another ceasefire was [announced](https://www.aljazeera.com/news/2022/9/15/armenia-azerbaijan-negotiate-another-ceasefire-yerevan) in the evening of 14th September 2022 by the secretary of Armenia’s Security Council. Within 2 days of clashes, [more than 100](https://www.aljazeera.com/news/2022/9/15/azerbaijan-says-71-soldiers-killed-in-armenia-border-clashes) Armenian and Azerbaijani soldiers were killed. +On 12th September 2022, fighting [erupted](https://www.bbc.com/news/world-europe-62888891) along the Azerbaijan-Armenia border between Azerbaijani and Armenian troops. The next day, Russia brokered a [ceasefire](https://www.bbc.com/news/world-europe-62828239), but it was immediately breached. Another ceasefire was [announced](https://www.aljazeera.com/news/2022/9/15/armenia-azerbaijan-negotiate-another-ceasefire-yerevan) in the evening of 14th September 2022 by the secretary of Armenia’s Security Council. Within 2 days of clashes, [more than 100](https://www.aljazeera.com/news/2022/9/15/azerbaijan-says-71-soldiers-killed-in-armenia-border-clashes) Armenian and Azerbaijani soldiers were killed. -This is the latest fighting [amid on-going clashes](https://globalvoices.org/2022/09/08/azerbaijan-armenia-and-tenuous-prospects-for-peace/) since [the 44-day war](https://www.bbc.com/news/world-europe-54324772) Armenia and Azerbaijan fought in 2020. +This is the latest fighting [amid on-going clashes](https://globalvoices.org/2022/09/08/azerbaijan-armenia-and-tenuous-prospects-for-peace/) since [the 44-day war](https://www.bbc.com/news/world-europe-54324772) Armenia and Azerbaijan fought in 2020. -In 1994, the Armenian military secured victory over Nagorno-Karabakh and the seven adjacent regions in the aftermath of the first Karabakh war. Since then, Nagorno-Karabakh has been under the control of its [ethnic Armenian population](https://globalvoices.org/2020/12/11/how-will-the-war-in-nagorno-karabakh-change-azerbaijan/) as a [self-declared state](https://globalvoices.org/2020/10/01/international-conflicts-are-also-about-names-the-case-of-nagorno-karabakh/). The two sides often blamed each other for violating the 1994 ceasefire agreement over the years with continued fighting. These clashes culminated in September 2020 during the second Karabakh war, with Azerbaijan [regaining](https://carnegieeurope.eu/2021/02/11/unfinished-business-in-armenia-azerbaijan-conflict-pub-83844) control over the previously occupied seven regions, as well as [one-third of Karabakh](https://carnegieeurope.eu/2021/02/11/unfinished-business-in-armenia-azerbaijan-conflict-pub-83844) itself. A ceasefire agreement signed on 10th November 2020 ended the hostilities temporarily. Despite both sides pledging to make progress towards a final peace deal and a [series of high level meetings](https://globalvoices.org/2022/05/25/armenian-and-azerbaijani-leaders-take-tentative-steps-toward-peace/) taking place since the second war, no significant progress on achieving a final peace treaty has been made.   +In 1994, the Armenian military secured victory over Nagorno-Karabakh and the seven adjacent regions in the aftermath of the first Karabakh war. Since then, Nagorno-Karabakh has been under the control of its [ethnic Armenian population](https://globalvoices.org/2020/12/11/how-will-the-war-in-nagorno-karabakh-change-azerbaijan/) as a [self-declared state](https://globalvoices.org/2020/10/01/international-conflicts-are-also-about-names-the-case-of-nagorno-karabakh/). The two sides often blamed each other for violating the 1994 ceasefire agreement over the years with continued fighting. These clashes culminated in September 2020 during the second Karabakh war, with Azerbaijan [regaining](https://carnegieeurope.eu/2021/02/11/unfinished-business-in-armenia-azerbaijan-conflict-pub-83844) control over the previously occupied seven regions, as well as [one-third of Karabakh](https://carnegieeurope.eu/2021/02/11/unfinished-business-in-armenia-azerbaijan-conflict-pub-83844) itself. A ceasefire agreement signed on 10th November 2020 ended the hostilities temporarily. Despite both sides pledging to make progress towards a final peace deal and a [series of high level meetings](https://globalvoices.org/2022/05/25/armenian-and-azerbaijani-leaders-take-tentative-steps-toward-peace/) taking place since the second war, no significant progress on achieving a final peace treaty has been made. -During the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), access to several social media platforms was [blocked](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war) in Azerbaijan. At the time (between September 2020 to November 2020), OONI data collected from Azerbaijan showed that several [social media websites](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-websites) (such as `facebook.com` and `twitter.com`) presented signs of blocking, while the testing of [WhatsApp](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT) and [Telegram](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp) presented signs of TLS level interference. +During the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), access to several social media platforms was [blocked](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war) in Azerbaijan. At the time (between September 2020 to November 2020), OONI data collected from Azerbaijan showed that several [social media websites](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-websites) (such as `facebook.com` and `twitter.com`) presented signs of blocking, while the testing of [WhatsApp](https://explorer.ooni.org/measurement/20201102T013112Z_whatsapp_AZ_39232_n1_jCcFoDIxyQlMrYtT) and [Telegram](https://explorer.ooni.org/measurement/20201102T183400Z_telegram_AZ_50274_n1_lngut2aTSYVobqHp) presented signs of TLS level interference. -During the latest clashes over the last days, OONI data shows that these social media platforms appear to be accessible in Azerbaijan, but [TikTok started to present signs of blocking](https://explorer.ooni.org/chart/mat?probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com&since=2022-08-16&until=2022-09-16&axis_x=measurement_start_day). On 14th September 2022, Azerbaijan’s state security services [announced](https://dtx.gov.az/az/news/1715.html) the temporary suspension of social media platform TikTok. According to the statement, the content circulating on the social media platform was revealing military secrets and forming wrong opinions in society. As a result, the State Security Service decided to temporarily block access to the platform. Similarly, OONI data shows that the testing of TikTok also started to present signs of [blocking in Armenia](https://explorer.ooni.org/chart/mat?probe_cc=AM&test_name=web_connectivity&domain=tiktok.com&since=2022-08-16&until=2022-09-16&axis_x=measurement_start_day) from 13th September 2022 onwards. +During the latest clashes over the last days, OONI data shows that these social media platforms appear to be accessible in Azerbaijan, but [TikTok started to present signs of blocking](https://explorer.ooni.org/chart/mat?probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com&since=2022-08-16&until=2022-09-16&axis_x=measurement_start_day). On 14th September 2022, Azerbaijan’s state security services [announced](https://dtx.gov.az/az/news/1715.html) the temporary suspension of social media platform TikTok. According to the statement, the content circulating on the social media platform was revealing military secrets and forming wrong opinions in society. As a result, the State Security Service decided to temporarily block access to the platform. Similarly, OONI data shows that the testing of TikTok also started to present signs of [blocking in Armenia](https://explorer.ooni.org/chart/mat?probe_cc=AM&test_name=web_connectivity&domain=tiktok.com&since=2022-08-16&until=2022-09-16&axis_x=measurement_start_day) from 13th September 2022 onwards. ## Methods Since 2012, OONI has developed free and open source software (called [OONI Probe](https://ooni.org/install/)) which is designed to [measure various forms of internet censorship](https://ooni.org/nettest/), including the blocking of websites and apps. Every month, OONI Probe is regularly run by volunteers in [around 170 countries](https://explorer.ooni.org/), and network measurements collected by OONI Probe users are automatically published as [open data in real-time](https://ooni.org/data). -This study involves the analysis of OONI measurements pertaining to the testing of TikTok, collected from local networks in [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM) over the last days. In particular, we analyzed [OONI Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity) pertaining to the testing of the TikTok homepage (`https://www.tiktok.com`), as well as several [other web resources that are needed for the TikTok web app and mobile app to function](https://github.com/citizenlab/test-lists/pull/1111) (hereafter referred to as “TikTok endpoints”). We thank a community member who kindly shared with us a list of TikTok endpoints, which were subsequently [added](https://github.com/citizenlab/test-lists/pulls?q=is%3Apr+is%3Aclosed) (to the [lists of URLs](https://ooni.org/support/faq#which-websites-will-i-test-for-censorship-with-ooni-probe) measured by OONI Probe users in [Azerbaijan](https://github.com/citizenlab/test-lists/pull/1111) and [Armenia](https://github.com/citizenlab/test-lists/pull/1112)) for testing. OONI’s [URL prioritization system](https://ooni.org/post/policy-url-prioritization/ooni-policy-url-prioritization.pdf) ensures that once URLs are merged into the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists), they are immediately prioritized for OONI Probe testing. As a result, measurements from the testing of TikTok endpoints were collected almost immediately from both Armenia and Azerbaijan (supporting this study). +This study involves the analysis of OONI measurements pertaining to the testing of TikTok, collected from local networks in [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM) over the last days. In particular, we analyzed [OONI Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity) pertaining to the testing of the TikTok homepage (`https://www.tiktok.com`), as well as several [other web resources that are needed for the TikTok web app and mobile app to function](https://github.com/citizenlab/test-lists/pull/1111) (hereafter referred to as “TikTok endpoints”). We thank a community member who kindly shared with us a list of TikTok endpoints, which were subsequently [added](https://github.com/citizenlab/test-lists/pulls?q=is%3Apr+is%3Aclosed) (to the [lists of URLs](https://ooni.org/support/faq#which-websites-will-i-test-for-censorship-with-ooni-probe) measured by OONI Probe users in [Azerbaijan](https://github.com/citizenlab/test-lists/pull/1111) and [Armenia](https://github.com/citizenlab/test-lists/pull/1112)) for testing. OONI’s [URL prioritization system](https://ooni.org/post/policy-url-prioritization/ooni-policy-url-prioritization.pdf) ensures that once URLs are merged into the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists), they are immediately prioritized for OONI Probe testing. As a result, measurements from the testing of TikTok endpoints were collected almost immediately from both Armenia and Azerbaijan (supporting this study). -OONI’s [Web Connectivity experiment](https://ooni.org/nettest/web-connectivity/) is designed to measure the accessibility of URLs by performing the following steps: +OONI’s [Web Connectivity experiment](https://ooni.org/nettest/web-connectivity/) is designed to measure the accessibility of URLs by performing the following steps: * Resolver identification * DNS lookup @@ -41,31 +41,31 @@ OONI’s [Web Connectivity experiment](https://ooni.org/nettest/web-connectivity * TLS handshake to the resolved IP addresses * HTTP(s) GET request -The above steps are automatically performed from both the local network of the user, and from a control vantage point. If the results from both networks are the same, the tested URL is annotated as `accessible`. If the results differ, the tested URL is annotated as `anomalous`, and the type of [anomaly](https://ooni.org/support/faq#what-do-you-mean-by-anomalies) is further characterized depending on the reason that caused the failure (for example, if the TCP connection fails, the measurement is annotated as a `TCP/IP anomaly`). +The above steps are automatically performed from both the local network of the user, and from a control vantage point. If the results from both networks are the same, the tested URL is annotated as `accessible`. If the results differ, the tested URL is annotated as `anomalous`, and the type of [anomaly](https://ooni.org/support/faq#what-do-you-mean-by-anomalies) is further characterized depending on the reason that caused the failure (for example, if the TCP connection fails, the measurement is annotated as a `TCP/IP anomaly`). -Each [Web Connectivity](https://ooni.org/nettest/web-connectivity/) measurement provides further network information (such as information pertaining to TLS handshakes) that helps with evaluating whether an anomalous measurement presents signs of blocking (or is a [false positive](https://ooni.org/support/faq#what-are-false-positives)). Based on our heuristics, we are able to automatically confirm the blocking of websites if a [block page](https://ooni.org/support/glossary#block-page) is served, or if DNS resolution returns an IP known to be associated with censorship (and such fingerprints have been added to our database). +Each [Web Connectivity](https://ooni.org/nettest/web-connectivity/) measurement provides further network information (such as information pertaining to TLS handshakes) that helps with evaluating whether an anomalous measurement presents signs of blocking (or is a [false positive](https://ooni.org/support/faq#what-are-false-positives)). Based on our heuristics, we are able to automatically confirm the blocking of websites if a [block page](https://ooni.org/support/glossary#block-page) is served, or if DNS resolution returns an IP known to be associated with censorship (and such fingerprints have been added to our database). -In the cases of Azerbaijan and Armenia, we have never come across block pages, limiting our ability to automatically confirm cases of blocking with confidence. Instead, ISPs in both Azerbaijan and Armenia appear to adopt different censorship techniques (as discussed, for example, in our [previous research](https://ooni.org/post/2021-azerbaijan/) examining blocks in Azerbaijan), requiring more manual analysis of measurements for the confirmation of blocks. This study focused in the case of Azerbaijan, we analyzed network measurement data provided under the [network_events](https://github.com/ooni/spec/blob/master/data-formats/df-008-netevents.md) keys of Web Connectivity measurements, which provided insight on TLS level interference of TikTok endpoints. In the case of Armenia, we analyzed both TLS handshakes and DNS query results. Another limitation to our study is the [limited OONI measurement coverage from Azerbaijan](https://explorer.ooni.org/country/AZ) (in comparison to [Armenia](https://explorer.ooni.org/country/AM) and [other countries](https://explorer.ooni.org/countries)). Generally, the larger the relevant measurement volume, the greater our ability in identifying and confirming blocks. +In the cases of Azerbaijan and Armenia, we have never come across block pages, limiting our ability to automatically confirm cases of blocking with confidence. Instead, ISPs in both Azerbaijan and Armenia appear to adopt different censorship techniques (as discussed, for example, in our [previous research](https://ooni.org/post/2021-azerbaijan/) examining blocks in Azerbaijan), requiring more manual analysis of measurements for the confirmation of blocks. This study focused in the case of Azerbaijan, we analyzed network measurement data provided under the [network_events](https://github.com/ooni/spec/blob/master/data-formats/df-008-netevents.md) keys of Web Connectivity measurements, which provided insight on TLS level interference of TikTok endpoints. In the case of Armenia, we analyzed both TLS handshakes and DNS query results. Another limitation to our study is the [limited OONI measurement coverage from Azerbaijan](https://explorer.ooni.org/country/AZ) (in comparison to [Armenia](https://explorer.ooni.org/country/AM) and [other countries](https://explorer.ooni.org/countries)). Generally, the larger the relevant measurement volume, the greater our ability in identifying and confirming blocks. ## Findings -OONI data shows that both [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM&test_name=web_connectivity&domain=www.tiktok.com) started blocking access to TikTok right after clashes erupted along their border. +OONI data shows that both [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM&test_name=web_connectivity&domain=www.tiktok.com) started blocking access to TikTok right after clashes erupted along their border. ### Blocking of TikTok in Azerbaijan -We analyzed [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) from 13th-15th September 2022 and noticed that many of them appear to time out when attempting to establish a TLS handshake to the target endpoint. +We analyzed [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) from 13th-15th September 2022 and noticed that many of them appear to time out when attempting to establish a TLS handshake to the target endpoint. -While many such measurements are annotated (on [OONI Explorer](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com)) as presenting DNS inconsistency, we were able to exclude that by validating that the returned IPs are able to complete a TLS handshake with a valid certificate for the target domain name. Moreover, when issuing an HTTP request with the appropriate headers, the response payload is consistent with the response from an endpoint served to a user in Europe. By inspecting the response header  and server certificate, it seems to be an Akamai cache. +While many such measurements are annotated (on [OONI Explorer](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com)) as presenting DNS inconsistency, we were able to exclude that by validating that the returned IPs are able to complete a TLS handshake with a valid certificate for the target domain name. Moreover, when issuing an HTTP request with the appropriate headers, the response payload is consistent with the response from an endpoint served to a user in Europe. By inspecting the response header and server certificate, it seems to be an Akamai cache. To evaluate the extent to which the blocking is happening and establish if it’s in fact caused by an intentional block, we analyzed the values of the `network_events` keys of [Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com). These keys contain very rich information on every read and write operation performed during a TLS handshake. -In the following charts, we consider a single “datapoint” as a single network event transcript (i.e. a TLS handshake operation towards a certain IP, port, server\_name tuple). When grouping the measurements by server\_name, we can see that most domains fail quite consistently, yet some of them don’t fail as often as others. +In the following charts, we consider a single “datapoint” as a single network event transcript (i.e. a TLS handshake operation towards a certain IP, port, server\_name tuple). When grouping the measurements by server_name, we can see that most domains fail quite consistently, yet some of them don’t fail as often as others. ![](images/image2.png) -Chart: OONI data on the testing of tiktok.com and TikTok endpoints in Azerbaijan between 13th-15th September 2022. +**Chart:** OONI data on the testing of tiktok.com and TikTok endpoints in Azerbaijan between 13th-15th September 2022. -As is evident from the above chart, the testing of `www.tiktok.com` and Tiktok endpoints presented a relatively large volume of anomalies between 13th to 15th September 2022, following the eruption of border clashes on 12th September 2022. It’s worth noting though that while www.[tiktok.com measurements](https://explorer.ooni.org/chart/mat?probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com&since=2022-08-17&until=2022-09-17&axis_x=measurement_start_day) (showing signs of blocking) were collected from Azerbaijan from 13th September onwards, TikTok endpoint measurements are only available from 14th September 2022 onwards (when they were [added](https://github.com/citizenlab/test-lists/pull/1111) to the test list for Azerbaijan). +As is evident from the above chart, the testing of `www.tiktok.com` and Tiktok endpoints presented a relatively large volume of anomalies between 13th to 15th September 2022, following the eruption of border clashes on 12th September 2022. It’s worth noting though that while www.[tiktok.com measurements](https://explorer.ooni.org/chart/mat?probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com&since=2022-08-17&until=2022-09-17&axis_x=measurement_start_day) (showing signs of blocking) were collected from Azerbaijan from 13th September onwards, TikTok endpoint measurements are only available from 14th September 2022 onwards (when they were [added](https://github.com/citizenlab/test-lists/pull/1111) to the test list for Azerbaijan). As a result, based on available OONI data, we can only confirm that the blocking of TikTok in Azerbaijan started the latest at the following times, on the following networks: @@ -77,37 +77,37 @@ The OONI Probe testing of `www.tiktok.com` on 5 networks in Azerbaijan between 1 ![](images/image1.png) -Chart: OONI Probe testing of `www.tiktok.com` on 5 networks in Azerbaijan between 13th-15th September 2022. +**Chart:** OONI Probe testing of `www.tiktok.com` on 5 networks in Azerbaijan between 13th-15th September 2022. -In the above cases, we can see that attempts to connect to `www.tiktok.com` failed due to a connection timeout. Upon inspecting [relevant network measurement data](https://explorer.ooni.org/search?since=2022-09-13&until=2022-09-16&failure=false&domain=www.tiktok.com&probe_cc=AZ), we see that this always happened after the first write operation (the one responsible for writing the ClientHello which contains the server\_name field (SNI)), which is consistent with the hypothesis that the block is some form of TLS based interference, possibly by looking at the SNI field of the ClientHello message. +In the above cases, we can see that attempts to connect to `www.tiktok.com` failed due to a connection timeout. Upon inspecting [relevant network measurement data](https://explorer.ooni.org/search?since=2022-09-13&until=2022-09-16&failure=false&domain=www.tiktok.com&probe_cc=AZ), we see that this always happened after the first write operation (the one responsible for writing the ClientHello which contains the server_name field (SNI)), which is consistent with the hypothesis that the block is some form of TLS based interference, possibly by looking at the SNI field of the ClientHello message. The blocking looks quite consistent on a per domain basis, which is compatible with the hypothesis that it’s some sort of SNI-based TLS level block. -To additionally support this claim, we notice that several different domain names resolve to the same IP address. When looking at the TLS reachability of the same IPs on a particular network, but with a different server\_name value, we can see (by reading the chart horizontally) that they are consistently reachable when a specific server\_name is provided, as illustrated below. +To additionally support this claim, we notice that several different domain names resolve to the same IP address. When looking at the TLS reachability of the same IPs on a particular network, but with a different server\_name value, we can see (by reading the chart horizontally) that they are consistently reachable when a specific server_name is provided, as illustrated below. ![](images/image5.png) -Chart: OONI data on TLS handshake results by target IP on network AS531712 in Azerbaijan on 15th September 2022. +**Chart:** OONI data on TLS handshake results by target IP on network AS531712 in Azerbaijan on 15th September 2022. -On the other hand, if you read the chart vertically you can see that the same IP is both reachable and unreachable depending on the domain\_name value. +On the other hand, if you read the chart vertically you can see that the same IP is both reachable and unreachable depending on the domain_name value. -This strongly suggests that the block is intentional and a consequence of filtering of those particular domain names. +This strongly suggests that the block is intentional and a consequence of filtering of those particular domain names. -When looking at the subset of domain names that are not consistently blocked, we can see that they are consistently blocked within the same network, which leads us to believe that there are different filtering policies deployed across networks. +When looking at the subset of domain names that are not consistently blocked, we can see that they are consistently blocked within the same network, which leads us to believe that there are different filtering policies deployed across networks. ![](images/image4.png) -Chart: OONI data providing a per-ASN comparison on the testing of `www.tiktok.com` and TikTok endpoint domains in Azerbaijan between 13th-15th September 2022. +**Chart:** OONI data providing a per-ASN comparison on the testing of `www.tiktok.com` and TikTok endpoint domains in Azerbaijan between 13th-15th September 2022. In summary, based on the analysis of 681 OONI metrics collected from 5 different networks in Azerbaijan, we conclude that: -* The main TikTok website (`www.tiktok.com`) and several endpoints essential to its functionality were blocked on at least 3 different networks (`AS29049`, `AS41997`, `AS31721`); -* On all networks where we identified blocking, it seems to be implemented by means of TLS level interference by dropping packets after noticing a disallowed server\_name; +* The main TikTok website (`www.tiktok.com`) and several endpoints essential to its functionality were blocked on at least 3 different networks (`AS29049`, `AS41997`, `AS31721`); +* On all networks where we identified blocking, it seems to be implemented by means of TLS level interference by dropping packets after noticing a disallowed server_name; * ISPs in Azerbaijan block TikTok based on different lists of TikTok endpoint domains, and there is some level of inconsistency. ### Blocking of TikTok in Armenia -OONI data shows the [blocking of TikTok in Armenia](https://explorer.ooni.org/chart/mat?probe_cc=AM&test_name=web_connectivity&domain=tiktok.com&since=2022-08-17&until=2022-09-17&axis_x=measurement_start_day) on 13th September 2022. Out of the 5 networks from which TikTok measurements were collected, we are able to establish that the blocking began on 2 of them on 13th September 2022. Specifically: +OONI data shows the [blocking of TikTok in Armenia](https://explorer.ooni.org/chart/mat?probe_cc=AM&test_name=web_connectivity&domain=tiktok.com&since=2022-08-17&until=2022-09-17&axis_x=measurement_start_day) on 13th September 2022. Out of the 5 networks from which TikTok measurements were collected, we are able to establish that the blocking began on 2 of them on 13th September 2022. Specifically: * [AS43733 2022-09-13 12:05:33](https://explorer.ooni.org/measurement/20220913T120506Z_webconnectivity_AM_43733_n1_ejea5bTX30wLekZe?input=https%3A%2F%2Ftiktok.com) * [AS44395 2022-09-13 12:03:50](https://explorer.ooni.org/measurement/20220913T120347Z_webconnectivity_AM_44395_n1_dMsbSChpjjLus2T5?input=https%3A%2F%2Ftiktok.com) @@ -116,7 +116,7 @@ The following chart shares OONI data on the testing of `www.tiktok.com` and `tik ![](images/image3.png) -Chart: OONI data on the testing of `tiktok.com` and `www.tiktok.com` on 5 networks in Armenia between 13th-15th September 2022. +**Chart:** OONI data on the testing of `tiktok.com` and `www.tiktok.com` on 5 networks in Armenia between 13th-15th September 2022. As is evident from the above charts, these ISPs in Armenia seem to employ a mix of different techniques to implement the blocking of TikTok. In the case of Ucom (AS44395), they appear to mostly return an NXDOMAIN error when the TikTok domain is queried. @@ -140,13 +140,13 @@ tiktok.com.                1800        IN        ;; XFR size: 5 records (messages 1, bytes 196) ``` -No other [tested TikTok endpoints](https://github.com/citizenlab/test-lists/pull/1112/files) allow zone transfers from the MTS nameserver, nor any other domain which we tested. This might be due to the fact that to implement the block the zone for tiktok.com was set up quickly and they forgot to disable zone transfers on it. +No other [tested TikTok endpoints](https://github.com/citizenlab/test-lists/pull/1112/files) allow zone transfers from the MTS nameserver, nor any other domain which we tested. This might be due to the fact that to implement the block the zone for tiktok.com was set up quickly and they forgot to disable zone transfers on it. -By focusing our attention on the Ucom network (AS44395), we can see that not all TikTok endpoints appear to be blocked. +By focusing our attention on the Ucom network (AS44395), we can see that not all TikTok endpoints appear to be blocked. ![](images/image6.png) -Chart: OONI data on the testing of TikTok endpoints on Ucom (AS44395) in Azerbaijan between 13th-16th September 2022. +**Chart:** OONI data on the testing of TikTok endpoints on Ucom (AS44395) in Azerbaijan between 13th-16th September 2022. Interestingly, starting from today (16th September 2022), the DNS response appears to be inconsistent and returns an IP address that has an expired letsencrypt certificate. When accessed, it will direct to: @@ -184,18 +184,18 @@ In all other cases of blocking, we notice that the HTTPS request times out, whic To summarize: * Armenia has been blocking TikTok since at least 13th September 2022 on at least two networks (AS44395, AS43733); -* On most networks, TikTok interference was observed at the DNS level by returning an NXDOMAIN error, an IP address associated with `www.google.com`, or a set of unrelated IP addresses which contain an expired certificate for the server\_name it.domain.name; +* On most networks, TikTok interference was observed at the DNS level by returning an NXDOMAIN error, an IP address associated with `www.google.com`, or a set of unrelated IP addresses which contain an expired certificate for the server_name it.domain.name; * Not all tested networks in Armenia implemented the TikTok block and censorship techniques varied from network to network; * Several endpoints used by the TikTok app and website are also blocked on at least one network in Armenia. ## Conclusion -Social media censorship often appears to be weaponized during conflicts and other moments of political turmoil around the world. By controlling the flow of information on social media platforms, governments may attempt to control the narrative surrounding political events. Over the last 6 years, OONI data has shown that [access to major social media platforms is commonly blocked](https://ooni.org/documents/2022-ooni-submission-ohchr-report-internet-shutdowns.pdf) during elections and protests around the world. Earlier this year, Russia [ramped up its censorship](https://ooni.org/post/2022-russia-blocks-amid-ru-ua-conflict/) following the war in Ukraine. +Social media censorship often appears to be weaponized during conflicts and other moments of political turmoil around the world. By controlling the flow of information on social media platforms, governments may attempt to control the narrative surrounding political events. Over the last 6 years, OONI data has shown that [access to major social media platforms is commonly blocked](https://ooni.org/documents/2022-ooni-submission-ohchr-report-internet-shutdowns.pdf) during elections and protests around the world. Earlier this year, Russia [ramped up its censorship](https://ooni.org/post/2022-russia-blocks-amid-ru-ua-conflict/) following the war in Ukraine. -Azerbaijan previously [blocked access to social media platforms during the 2020 Nagorno-Karabakh war](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war). In similar vein, both [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM&test_name=web_connectivity&domain=tiktok.com) blocked access to TikTok during the border clashes over the last days. OONI data shows that Azerbaijan blocked TikTok by means of TLS level interference, while Armenia implemented both DNS level blocks and TLS level interference. +Azerbaijan previously [blocked access to social media platforms during the 2020 Nagorno-Karabakh war](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war). In similar vein, both [Azerbaijan](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AZ&test_name=web_connectivity&domain=www.tiktok.com) and [Armenia](https://explorer.ooni.org/search?since=2022-08-17&until=2022-09-17&failure=false&probe_cc=AM&test_name=web_connectivity&domain=tiktok.com) blocked access to TikTok during the border clashes over the last days. OONI data shows that Azerbaijan blocked TikTok by means of TLS level interference, while Armenia implemented both DNS level blocks and TLS level interference. You can continue to monitor the reachability of TikTok (and other platforms) in Armenia and Azerbaijan through [OONI data](https://explorer.ooni.org/chart/mat). ## Acknowledgements -We thank [OONI Probe](https://ooni.org/install/) users in Armenia and Azerbaijan who contributed measurements, supporting this study. +We thank [OONI Probe](https://ooni.org/install/) users in Armenia and Azerbaijan who contributed measurements, supporting this study. diff --git a/content/post/2022-iran-blocks-social-media-mahsa-amini-protests/index.md b/content/post/2022-iran-blocks-social-media-mahsa-amini-protests/index.md index 3577af808..f64a8d3ac 100644 --- a/content/post/2022-iran-blocks-social-media-mahsa-amini-protests/index.md +++ b/content/post/2022-iran-blocks-social-media-mahsa-amini-protests/index.md @@ -2,7 +2,7 @@ title: "Iran blocks social media, app stores and encrypted DNS amid Mahsa Amini protests" author: "Simone Basso (OONI), Maria Xynou (OONI), Arturo Filastò (OONI), Amanda Meng (IODA - Georgia Tech)" date: "2022-09-25" -tags: ["iran", "censorship", "country-ir", "mahsaamini", "protests", "mahsa", "amini"] +tags: ["iran", "censorship", "country-ir", "mahsaamini", "protests", "mahsa", "amini", "theme-social_media", "theme-im"] categories: ["report"] --- diff --git a/content/post/2022-iran-technical-multistakeholder-report/index.md b/content/post/2022-iran-technical-multistakeholder-report/index.md index 466fcf91c..a87e3fc65 100644 --- a/content/post/2022-iran-technical-multistakeholder-report/index.md +++ b/content/post/2022-iran-technical-multistakeholder-report/index.md @@ -2,7 +2,7 @@ title: "Technical multi-stakeholder report on Internet shutdowns: The case of Iran amid autumn 2022 protests" date: "2022-11-29" author: "OONI, IODA, M-Lab, Cloudflare, Kentik, Censored Planet, ISOC, Article19" -tags: ["iran", "censorship", "country-ir", "mahsaamini", "protests", "shutdown"] +tags: ["iran", "censorship", "country-ir", "mahsaamini", "protests", "shutdown", "theme-im", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2022-ooni-submission-ohchr-report-internet-shutdowns/index.md b/content/post/2022-ooni-submission-ohchr-report-internet-shutdowns/index.md index ca3fbe593..e02b75907 100644 --- a/content/post/2022-ooni-submission-ohchr-report-internet-shutdowns/index.md +++ b/content/post/2022-ooni-submission-ohchr-report-internet-shutdowns/index.md @@ -3,7 +3,7 @@ title: "OONI’s submission for the OHCHR report on internet shutdowns and human description: "This post shares OONI's submission in support of the OHCHR report on internet shutdowns and human rights to the 50th session of the Human Rights Council June 2022." author: "Maria Xynou" date: "2022-06-15" -tags: ["ooni", "ohchr", "submission", "ooni-data"] +tags: ["ooni", "ohchr", "submission", "ooni-data", "theme-social_media"] categories: ["blog"] --- diff --git a/content/post/2022-russia-blocks-amid-ru-ua-conflict/index.md b/content/post/2022-russia-blocks-amid-ru-ua-conflict/index.md index 2f42d2ddf..1428d3476 100644 --- a/content/post/2022-russia-blocks-amid-ru-ua-conflict/index.md +++ b/content/post/2022-russia-blocks-amid-ru-ua-conflict/index.md @@ -3,7 +3,7 @@ title: "New blocks emerge in Russia amid war in Ukraine: An OONI network measure description: "This research report shared OONI data analysis on recent blocks that have emerged in Russia(and in the EU) amid the war in Ukraine." author: "Maria Xynou, Arturo Filastò" date: "2022-03-07" -tags: ["russia", "censorship", "country-ru", "country-gr", "country-es", "country-de", "country-fr", "country-ro", "country-pl", "country-cz", "country-ua"] +tags: ["russia", "censorship", "country-ru", "country-gr", "country-es", "country-de", "country-fr", "country-ro", "country-pl", "country-cz", "country-ua", "theme-news_media", "theme-social_media"] categories: ["report"] --- @@ -11,7 +11,7 @@ categories: ["report"] Information controls are known to occur during conflicts, and that’s exactly what we’re seeing in Russia following the recent invasion of Ukraine on 24th February 2022. -In recent days, [OONI network measurement data](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU) collected from Russia shows that many Internet Service Providers (ISPs) have started [blocking](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies) access to several news media websites, as well as to a [website](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies&domain=200rf.com) (200rf.com) that shares information about captured and killed Russian soldiers in Ukraine. OONI data also shows that Russian ISPs started [throttling access to Twitter](https://explorer.ooni.org/search?since=2022-02-26&until=2022-03-04&failure=false&domain=twitter.com&probe_cc=RU) on 26th February 2022, and switched to [blocking](https://explorer.ooni.org/measurement/20220306T202033Z_webconnectivity_RU_51604_n1_oVoIGdnUmlxCWmyR?input=https%3A%2F%2Ftwitter.com%2F) it by 4th March 2022 – at which point, they also started [blocking access to Facebook](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies). Censorship in Russia is generally implemented in a [decentralized](https://www.ndss-symposium.org/ndss-paper/decentralized-control-a-case-study-of-russia/) way, as blocks are not observed on all networks, while ISPs adopt a variety of different censorship techniques. +In recent days, [OONI network measurement data](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU) collected from Russia shows that many Internet Service Providers (ISPs) have started [blocking](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies) access to several news media websites, as well as to a [website](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies&domain=200rf.com) (200rf.com) that shares information about captured and killed Russian soldiers in Ukraine. OONI data also shows that Russian ISPs started [throttling access to Twitter](https://explorer.ooni.org/search?since=2022-02-26&until=2022-03-04&failure=false&domain=twitter.com&probe_cc=RU) on 26th February 2022, and switched to [blocking](https://explorer.ooni.org/measurement/20220306T202033Z_webconnectivity_RU_51604_n1_oVoIGdnUmlxCWmyR?input=https%3A%2F%2Ftwitter.com%2F) it by 4th March 2022 – at which point, they also started [blocking access to Facebook](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies). Censorship in Russia is generally implemented in a [decentralized](https://www.ndss-symposium.org/ndss-paper/decentralized-control-a-case-study-of-russia/) way, as blocks are not observed on all networks, while ISPs adopt a variety of different censorship techniques. In this report, we share an in-depth analysis of new censorship events that have recently emerged in Russia based on [OONI network measurement data](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies). @@ -19,59 +19,59 @@ In this report, we share an in-depth analysis of new censorship events that have ## Key Findings -* **Media censorship.** Russia recently started blocking access to independent Russian news media websites (such as [Dozhd](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=tvrain.ru&probe_cc=RU&only=anomalies) and [New Times](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=newtimes.ru&probe_cc=RU&only=anomalies)) and foreign news media websites (such as [BBC](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.bbc.com&probe_cc=RU&only=anomalies), [Deutsche Welle](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.dw.com&probe_cc=RU&only=anomalies), [Russian version of Voice of America](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU&only=anomalies), and [Russian service of Radio Liberty](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.svoboda.org&probe_cc=RU&only=anomalies)). -* **Blocking of a website about captured and killed Russian soldiers.** Russia [blocked access to 200rf.com](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-06&failure=false&domain=200rf.com&probe_cc=RU), which was created by representatives of the Ministry of Internal Affairs of Ukraine to [share information about captured and killed Russian soldiers](https://200rf.com/) in Ukraine. -* **Centralized throttling of Twitter.** OONI data suggests that access to [twitter.com was throttled](https://explorer.ooni.org/search?since=2022-02-26&until=2022-03-04&failure=false&domain=twitter.com&probe_cc=RU) in Russia between 26th February 2022 to 4th March 2022. As the throttling of twitter.com seems to have stopped across all ISPs in Russia at the same time (~08:00 on 4th March 2022 UTC), it appears to have been centralized. Moreover, interference to twitter.com appears to have changed from throttling to blocking (through the injection of a RST packet) on 4th March 2022. -* **Blocking of Facebook and Twitter.** OONI data shows that access to [facebook.com](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies) and [twitter.com was blocked](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU) in Russia by 4th March 2022 (primarily through the injection of RST packets). -* **Decentralized censorship.** Every Internet Service Provider (ISP) in Russia is responsible for implementing government-mandated blocks independently. As a result, we observe variance in how internet censorship is implemented across Russia, as blocks are not implemented on all networks in the country, and different ISPs adopt different censorship techniques. Some ISPs implement blocks through the use of multiple techniques at the same time, making circumvention harder. -* **Different censorship techniques.** To block websites, OONI data shows that Russian ISPs adopt the following censorship techniques (beyond throttling): +* **Media censorship.** Russia recently started blocking access to independent Russian news media websites (such as [Dozhd](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=tvrain.ru&probe_cc=RU&only=anomalies) and [New Times](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=newtimes.ru&probe_cc=RU&only=anomalies)) and foreign news media websites (such as [BBC](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.bbc.com&probe_cc=RU&only=anomalies), [Deutsche Welle](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.dw.com&probe_cc=RU&only=anomalies), [Russian version of Voice of America](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU&only=anomalies), and [Russian service of Radio Liberty](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.svoboda.org&probe_cc=RU&only=anomalies)). +* **Blocking of a website about captured and killed Russian soldiers.** Russia [blocked access to 200rf.com](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-06&failure=false&domain=200rf.com&probe_cc=RU), which was created by representatives of the Ministry of Internal Affairs of Ukraine to [share information about captured and killed Russian soldiers](https://200rf.com/) in Ukraine. +* **Centralized throttling of Twitter.** OONI data suggests that access to [twitter.com was throttled](https://explorer.ooni.org/search?since=2022-02-26&until=2022-03-04&failure=false&domain=twitter.com&probe_cc=RU) in Russia between 26th February 2022 to 4th March 2022. As the throttling of twitter.com seems to have stopped across all ISPs in Russia at the same time (~08:00 on 4th March 2022 UTC), it appears to have been centralized. Moreover, interference to twitter.com appears to have changed from throttling to blocking (through the injection of a RST packet) on 4th March 2022. +* **Blocking of Facebook and Twitter.** OONI data shows that access to [facebook.com](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies) and [twitter.com was blocked](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU) in Russia by 4th March 2022 (primarily through the injection of RST packets). +* **Decentralized censorship.** Every Internet Service Provider (ISP) in Russia is responsible for implementing government-mandated blocks independently. As a result, we observe variance in how internet censorship is implemented across Russia, as blocks are not implemented on all networks in the country, and different ISPs adopt different censorship techniques. Some ISPs implement blocks through the use of multiple techniques at the same time, making circumvention harder. +* **Different censorship techniques.** To block websites, OONI data shows that Russian ISPs adopt the following censorship techniques (beyond throttling): * [DNS manipulation, redirecting in some cases to blockpages](https://explorer.ooni.org/measurement/20220305T042324Z_webconnectivity_RU_35807_n1_7gOd4SMmKjQ3mmYn?input=https%3A%2F%2Fwww.bbc.com%2Frussian%2F) * [HTTP man-in-the-middle, serving blockpages](https://explorer.ooni.org/measurement/20220304T214241Z_webconnectivity_RU_12389_n1_UeXMqEpughRTTzLU?input=http%3A%2F%2Fwww.bbc.com%2Fnews) * [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220304T050209Z_webconnectivity_RU_8359_n1_Ajrlzw7jakFRM3pM?input=https%3A%2F%2Fwww.bbc.com%2F) - * [Injection of a RST packet after the ClientHello during the TLS handshake](https://explorer.ooni.org/measurement/20220305T090244Z_webconnectivity_RU_8732_n1_oCVFwxN1oAYddBoX?input=https%3A%2F%2Fwww.bbc.com%2Frussian%2F) (most prevalent) + * [Injection of a RST packet after the ClientHello during the TLS handshake](https://explorer.ooni.org/measurement/20220305T090244Z_webconnectivity_RU_8732_n1_oCVFwxN1oAYddBoX?input=https%3A%2F%2Fwww.bbc.com%2Frussian%2F) (most prevalent) * [Timing out the session after the ClientHello during the TLS handshake](https://explorer.ooni.org/measurement/20220304T080535Z_webconnectivity_RU_16345_n1_3v6HMrLt0KTMXcpm?input=https%3A%2F%2Fwww.bbc.com%2F) * [Closing the connection after the ClientHello during the TLS handshake](https://explorer.ooni.org/measurement/20220304T034651Z_webconnectivity_RU_31163_n1_asVR1x6WcRk5AdL0?input=https%3A%2F%2Fwww.bbc.com%2F) ## Background -On 24th February 2022, [Russia invaded Ukraine](https://www.dw.com/en/russias-invasion-of-ukraine-2022/t-60931789) and the military operation is currently ongoing. The conflict between the two nations began 8 years ago (February 2014), when Russia invaded and [annexed the Crimean Peninsula](https://www.bbc.com/news/world-europe-26644082) from Ukraine. This was followed by the [war in the eastern Donbas region](https://www.bbc.com/news/world-europe-59599066) of Ukraine (which has since been ongoing), where Russian separatist groups have been backed by the Russian military. Over the years, the conflict has also involved cyberwarfare, with a series of [cyberattacks](https://en.wikipedia.org/wiki/Russian%E2%80%93Ukrainian_cyberwarfare) reportedly launched by both Russia and Ukraine. +On 24th February 2022, [Russia invaded Ukraine](https://www.dw.com/en/russias-invasion-of-ukraine-2022/t-60931789) and the military operation is currently ongoing. The conflict between the two nations began 8 years ago (February 2014), when Russia invaded and [annexed the Crimean Peninsula](https://www.bbc.com/news/world-europe-26644082) from Ukraine. This was followed by the [war in the eastern Donbas region](https://www.bbc.com/news/world-europe-59599066) of Ukraine (which has since been ongoing), where Russian separatist groups have been backed by the Russian military. Over the years, the conflict has also involved cyberwarfare, with a series of [cyberattacks](https://en.wikipedia.org/wiki/Russian%E2%80%93Ukrainian_cyberwarfare) reportedly launched by both Russia and Ukraine. -Leading up to and following the current full-scale invasion of Ukraine, DDoS attacks have been launched against Ukrainian government and banking websites, which US officials [attributed](https://www.wired.com/story/ukraine-ddos-russia-crypo-roblox-security-news/) to Russia, but the Russian government [denied](https://www.cnbc.com/2022/02/23/cyberattack-hits-ukrainian-banks-and-government-websites.html) involvement. Russia was also accused of targeting the machines of multiple organizations in Ukraine with [data-wiper malware](https://www.theguardian.com/world/2022/feb/24/russia-unleashed-data-wiper-virus-on-ukraine-say-cyber-experts). Meanwhile, the Anonymous hacker collective [declared cyberwar](https://www.theguardian.com/world/2022/feb/27/anonymous-the-hacker-collective-that-has-declared-cyberwar-on-russia) against the Russian government, claiming credit for hacking the Russian Ministry of Defense database. They also reportedly hacked Russian state TV channels to post pro-Ukraine content and [taking down](https://twitter.com/YourAnonTV/status/1496968999900377088) the website (rt.com) of the Russian state news website RT. Amid the war, Ukraine has experienced [internet outages affecting several networks](https://ioda.inetintel.cc.gatech.edu/country/UA?from=1645315200&until=1646351999) over the last week, but overall, the country’s [internet infrastructure appears to continue to function](https://labs.ripe.net/author/emileaben/the-ukrainian-internet/) despite the conflict.   +Leading up to and following the current full-scale invasion of Ukraine, DDoS attacks have been launched against Ukrainian government and banking websites, which US officials [attributed](https://www.wired.com/story/ukraine-ddos-russia-crypo-roblox-security-news/) to Russia, but the Russian government [denied](https://www.cnbc.com/2022/02/23/cyberattack-hits-ukrainian-banks-and-government-websites.html) involvement. Russia was also accused of targeting the machines of multiple organizations in Ukraine with [data-wiper malware](https://www.theguardian.com/world/2022/feb/24/russia-unleashed-data-wiper-virus-on-ukraine-say-cyber-experts). Meanwhile, the Anonymous hacker collective [declared cyberwar](https://www.theguardian.com/world/2022/feb/27/anonymous-the-hacker-collective-that-has-declared-cyberwar-on-russia) against the Russian government, claiming credit for hacking the Russian Ministry of Defense database. They also reportedly hacked Russian state TV channels to post pro-Ukraine content and [taking down](https://twitter.com/YourAnonTV/status/1496968999900377088) the website (rt.com) of the Russian state news website RT. Amid the war, Ukraine has experienced [internet outages affecting several networks](https://ioda.inetintel.cc.gatech.edu/country/UA?from=1645315200&until=1646351999) over the last week, but overall, the country’s [internet infrastructure appears to continue to function](https://labs.ripe.net/author/emileaben/the-ukrainian-internet/) despite the conflict. -Over the years, both Russia and Ukraine have engaged in geopolitical censorship, as [Russia has been blocking Ukrainian websites](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&probe_cc=RU&only=confirmed) (such as [uapolitics.com](https://explorer.ooni.org/measurement/20220209T105254Z_webconnectivity_RU_34456_n1_KKpP9t7GsvOoSpso?input=http%3A%2F%2Fuapolitics.com%2F), [24tv.ua](https://explorer.ooni.org/measurement/20220205T225542Z_webconnectivity_RU_51813_n1_tvDoPQfDxPE9RLGu?input=https%3A%2F%2F24tv.ua%2F), [donbass.ua](https://explorer.ooni.org/measurement/20211202T235114Z_webconnectivity_RU_51570_n1_yvzXZ3JcMTOzGuKY?input=http%3A%2F%2Fdonbass.ua%2F)), while [Ukraine has been blocking Russian websites](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&probe_cc=UA&only=confirmed) (such as [mail.ru](https://explorer.ooni.org/measurement/20220301T043010Z_webconnectivity_UA_35362_n1_5ncpRAizRmdWcaXq?input=http%3A%2F%2Fmail.ru%2F), [rt.com](https://explorer.ooni.org/measurement/20220302T043733Z_webconnectivity_UA_13188_n1_1NoiSEKgixSqo1OU?input=http%3A%2F%2Frt.com%2F), [gazeta.ru](https://explorer.ooni.org/measurement/20220302T062501Z_webconnectivity_UA_13188_n1_cz1btWU7agooErpV?input=http%3A%2F%2Fgazeta.ru%2F)). Russia though has been implementing more widespread internet censorship, also [blocking numerous LGBTIQ websites](https://ooni.org/post/2021-no-access-lgbtiq-website-censorship-six-countries/), a variety of websites that express political criticism (such as [ej.ru](https://explorer.ooni.org/measurement/20220303T142326Z_webconnectivity_RU_8402_n1_tZlHfk2hR4eLRhK6?input=http%3A%2F%2Fwww.ej.ru%2F), [imrussia.org](https://explorer.ooni.org/measurement/20220302T211436Z_webconnectivity_RU_51813_n1_DdYZMJBi9cKyvb7I?input=https%3A%2F%2Fimrussia.org%2Fru%2F), [kasparov.ru](https://explorer.ooni.org/measurement/20220303T050847Z_webconnectivity_RU_8790_n1_bua5WgKgLMSiTCqQ?input=http%3A%2F%2Fkasparov.ru%2F), [fbk.info](https://explorer.ooni.org/measurement/20220303T010633Z_webconnectivity_RU_51813_n1_woCOU3fsT8qop47t?input=https%3A%2F%2Ffbk.info%2F)), as well as the [website of opposition leader Alexei Navalny](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&probe_cc=RU&domain=navalny.com) (which OONI data suggests has been [blocked since 26th July 2021](https://explorer.ooni.org/search?since=2021-07-01&until=2021-07-31&failure=false&probe_cc=RU&domain=navalny.com)).   +Over the years, both Russia and Ukraine have engaged in geopolitical censorship, as [Russia has been blocking Ukrainian websites](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&probe_cc=RU&only=confirmed) (such as [uapolitics.com](https://explorer.ooni.org/measurement/20220209T105254Z_webconnectivity_RU_34456_n1_KKpP9t7GsvOoSpso?input=http%3A%2F%2Fuapolitics.com%2F), [24tv.ua](https://explorer.ooni.org/measurement/20220205T225542Z_webconnectivity_RU_51813_n1_tvDoPQfDxPE9RLGu?input=https%3A%2F%2F24tv.ua%2F), [donbass.ua](https://explorer.ooni.org/measurement/20211202T235114Z_webconnectivity_RU_51570_n1_yvzXZ3JcMTOzGuKY?input=http%3A%2F%2Fdonbass.ua%2F)), while [Ukraine has been blocking Russian websites](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&probe_cc=UA&only=confirmed) (such as [mail.ru](https://explorer.ooni.org/measurement/20220301T043010Z_webconnectivity_UA_35362_n1_5ncpRAizRmdWcaXq?input=http%3A%2F%2Fmail.ru%2F), [rt.com](https://explorer.ooni.org/measurement/20220302T043733Z_webconnectivity_UA_13188_n1_1NoiSEKgixSqo1OU?input=http%3A%2F%2Frt.com%2F), [gazeta.ru](https://explorer.ooni.org/measurement/20220302T062501Z_webconnectivity_UA_13188_n1_cz1btWU7agooErpV?input=http%3A%2F%2Fgazeta.ru%2F)). Russia though has been implementing more widespread internet censorship, also [blocking numerous LGBTIQ websites](https://ooni.org/post/2021-no-access-lgbtiq-website-censorship-six-countries/), a variety of websites that express political criticism (such as [ej.ru](https://explorer.ooni.org/measurement/20220303T142326Z_webconnectivity_RU_8402_n1_tZlHfk2hR4eLRhK6?input=http%3A%2F%2Fwww.ej.ru%2F), [imrussia.org](https://explorer.ooni.org/measurement/20220302T211436Z_webconnectivity_RU_51813_n1_DdYZMJBi9cKyvb7I?input=https%3A%2F%2Fimrussia.org%2Fru%2F), [kasparov.ru](https://explorer.ooni.org/measurement/20220303T050847Z_webconnectivity_RU_8790_n1_bua5WgKgLMSiTCqQ?input=http%3A%2F%2Fkasparov.ru%2F), [fbk.info](https://explorer.ooni.org/measurement/20220303T010633Z_webconnectivity_RU_51813_n1_woCOU3fsT8qop47t?input=https%3A%2F%2Ffbk.info%2F)), as well as the [website of opposition leader Alexei Navalny](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&probe_cc=RU&domain=navalny.com) (which OONI data suggests has been [blocked since 26th July 2021](https://explorer.ooni.org/search?since=2021-07-01&until=2021-07-31&failure=false&probe_cc=RU&domain=navalny.com)). Following the recent Russian invasion of Ukraine, we started to hear of some new censorship events in Russia, which we corroborate with OONI data and share in this report. ## Methods -Since 2012, OONI has developed [free and open source software](https://github.com/ooni/), called [OONI Probe](https://ooni.org/install/), designed to [measure various forms of internet censorship](https://ooni.org/nettest/). OONI Probe is run by volunteers in around 160 countries every month, and their test results are automatically [published by OONI as open data in real-time](https://ooni.org/data/). More than [675 million network measurements](https://explorer.ooni.org/) have been collected and published from 23,000 networks in 240 countries and territories over the last decade. +Since 2012, OONI has developed [free and open source software](https://github.com/ooni/), called [OONI Probe](https://ooni.org/install/), designed to [measure various forms of internet censorship](https://ooni.org/nettest/). OONI Probe is run by volunteers in around 160 countries every month, and their test results are automatically [published by OONI as open data in real-time](https://ooni.org/data/). More than [675 million network measurements](https://explorer.ooni.org/) have been collected and published from 23,000 networks in 240 countries and territories over the last decade. -OONI Probe users in Russia, in particular, regularly contribute a large volume of measurements, having contributed [154,672,256 measurements from 2,600 local networks in Russia](https://explorer.ooni.org/country/RU) over the years. Every day, new measurements are collected from Russia and [openly published](https://explorer.ooni.org/search?since=2022-02-02&until=2022-03-05&failure=false&probe_cc=RU) in real-time. We base the analysis of this research on a subset of these measurements collected from Russia (and several European countries) over the last month. +OONI Probe users in Russia, in particular, regularly contribute a large volume of measurements, having contributed [154,672,256 measurements from 2,600 local networks in Russia](https://explorer.ooni.org/country/RU) over the years. Every day, new measurements are collected from Russia and [openly published](https://explorer.ooni.org/search?since=2022-02-02&until=2022-03-05&failure=false&probe_cc=RU) in real-time. We base the analysis of this research on a subset of these measurements collected from Russia (and several European countries) over the last month. -More specifically, this report involves an analysis of measurements collected from [OONI’s Web Connectivity test](https://ooni.org/nettest/web-connectivity/), which is designed to measure the blocking of websites (these websites are publicly hosted on the [Citizen Lab test list Github repository](https://github.com/citizenlab/test-lists/tree/master/lists)). This test measures the accessibility of websites by attempting to perform a [DNS lookup](https://ooni.org/support/glossary/#dns-lookup), TCP/IP connection, and [HTTP GET request](https://ooni.org/support/glossary/#http-request) from two vantage points: (1) the local vantage point of the user and (2) a control network (non-censored network). The results from both networks are automatically compared and if they match, the tested URL is annotated as “accessible” (if the testing succeeds from the control vantage point). If the results differ, the tested URL is annotated as “[anomalous](https://ooni.org/support/faq/#what-do-you-mean-by-anomalies)”, which may provide a signal of potential blocking. +More specifically, this report involves an analysis of measurements collected from [OONI’s Web Connectivity test](https://ooni.org/nettest/web-connectivity/), which is designed to measure the blocking of websites (these websites are publicly hosted on the [Citizen Lab test list Github repository](https://github.com/citizenlab/test-lists/tree/master/lists)). This test measures the accessibility of websites by attempting to perform a [DNS lookup](https://ooni.org/support/glossary/#dns-lookup), TCP/IP connection, and [HTTP GET request](https://ooni.org/support/glossary/#http-request) from two vantage points: (1) the local vantage point of the user and (2) a control network (non-censored network). The results from both networks are automatically compared and if they match, the tested URL is annotated as “accessible” (if the testing succeeds from the control vantage point). If the results differ, the tested URL is annotated as “[anomalous](https://ooni.org/support/faq/#what-do-you-mean-by-anomalies)”, which may provide a signal of potential blocking. -Depending on why the anomaly emerges, the anomalous measurement is automatically annotated as a `DNS`, `TCP/IP`, `HTTP diff`, or `HTTP failure` anomaly. For example, if the DNS lookup resolves to an IP address which differs from that resolved from the control vantage point, the measurement is annotated as a “DNS anomaly”, which may be a sign of [DNS tampering](https://ooni.org/support/glossary/#dns-tampering). +Depending on why the anomaly emerges, the anomalous measurement is automatically annotated as a `DNS`, `TCP/IP`, `HTTP diff`, or `HTTP failure` anomaly. For example, if the DNS lookup resolves to an IP address which differs from that resolved from the control vantage point, the measurement is annotated as a “DNS anomaly”, which may be a sign of [DNS tampering](https://ooni.org/support/glossary/#dns-tampering). -However, [false positives](https://ooni.org/support/faq/#what-are-false-positives) can occur, which is why we look at anomalous measurements in aggregate in order to determine if a tested URL consistently presents a large volume of anomalous measurements (in comparison to successful measurements) on a tested network. If a tested URL presents a large volume of anomalies, it may provide a stronger signal of potential blocking. If the types of anomalies are consistent (for example, always presenting DNS anomalies on a tested network), they offer an even stronger signal of potential censorship (since they suggest the use of a specific censorship technique, such as [DNS hijacking](https://ooni.org/support/glossary/#dns-hijacking)). But beyond aggregating anomalous measurements, we also analyze the raw data pertaining to anomalous measurements in order to identify the specific errors that occurred as part of the testing, offering insight into how a tested URL is potentially blocked. +However, [false positives](https://ooni.org/support/faq/#what-are-false-positives) can occur, which is why we look at anomalous measurements in aggregate in order to determine if a tested URL consistently presents a large volume of anomalous measurements (in comparison to successful measurements) on a tested network. If a tested URL presents a large volume of anomalies, it may provide a stronger signal of potential blocking. If the types of anomalies are consistent (for example, always presenting DNS anomalies on a tested network), they offer an even stronger signal of potential censorship (since they suggest the use of a specific censorship technique, such as [DNS hijacking](https://ooni.org/support/glossary/#dns-hijacking)). But beyond aggregating anomalous measurements, we also analyze the raw data pertaining to anomalous measurements in order to identify the specific errors that occurred as part of the testing, offering insight into how a tested URL is potentially blocked. -Based on our current heuristics, we automatically confirm the blocking of websites when a [block page](https://ooni.org/support/glossary/#block-page) is served and we have added the fingerprint of that blockpage to our database. We also automatically confirm the blocking of websites based on DNS answers containing IP addresses that are known to be associated with implementing internet censorship. For other forms of censorship, we analyze OONI data in order to aggregate anomalous measurements and identify why and how those anomalies occur, offering insight into additional cases of potential blocking. +Based on our current heuristics, we automatically confirm the blocking of websites when a [block page](https://ooni.org/support/glossary/#block-page) is served and we have added the fingerprint of that blockpage to our database. We also automatically confirm the blocking of websites based on DNS answers containing IP addresses that are known to be associated with implementing internet censorship. For other forms of censorship, we analyze OONI data in order to aggregate anomalous measurements and identify why and how those anomalies occur, offering insight into additional cases of potential blocking. -To enable the internet freedom community to benefit from our analysis (and to support rapid response efforts), we have been building a [Measurement Aggregation Toolkit (MAT)](https://explorer.ooni.org/experimental/mat) which enables users to produce custom charts based on aggregate views of OONI data. We use charts generated from the MAT in this report. +To enable the internet freedom community to benefit from our analysis (and to support rapid response efforts), we have been building a [Measurement Aggregation Toolkit (MAT)](https://explorer.ooni.org/experimental/mat) which enables users to produce custom charts based on aggregate views of OONI data. We use charts generated from the MAT in this report. -While OONI does not currently have a methodology specifically designed to automatically measure throttling, we are able to infer potential throttling by analyzing the `network_event` keys of the raw data in [Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-02-02&until=2022-03-05&failure=false&probe_cc=RU&test_name=web_connectivity). We share the findings of our analysis in the following sections of this report. +While OONI does not currently have a methodology specifically designed to automatically measure throttling, we are able to infer potential throttling by analyzing the `network_event` keys of the raw data in [Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-02-02&until=2022-03-05&failure=false&probe_cc=RU&test_name=web_connectivity). We share the findings of our analysis in the following sections of this report. ## Findings Following the invasion of Ukraine, OONI data shows that Russia started: * Blocking several [news media websites](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies) -* Blocking a [website](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies&domain=200rf.com) (200rf.com) that shares information about captured and killed Russian soldiers -* [Centralized throttling of Twitter](https://explorer.ooni.org/search?since=2022-02-26&until=2022-03-04&failure=false&domain=twitter.com&probe_cc=RU) (between 26th February 2022 to 4th March 2022) -* Blocking [Facebook](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies) and [Twitter](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU) (from 4th March 2022 onwards)  +* Blocking a [website](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity&only=anomalies&domain=200rf.com) (200rf.com) that shares information about captured and killed Russian soldiers +* [Centralized throttling of Twitter](https://explorer.ooni.org/search?since=2022-02-26&until=2022-03-04&failure=false&domain=twitter.com&probe_cc=RU) (between 26th February 2022 to 4th March 2022) +* Blocking [Facebook](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies) and [Twitter](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU) (from 4th March 2022 onwards) But these censorship events are not present on all networks in Russia. -Russia implements internet censorship in a [decentralized](https://www.ndss-symposium.org/ndss-paper/decentralized-control-a-case-study-of-russia/) manner, as every Internet Service Provider (ISP) in Russia is responsible for implementing government-mandated blocks independently. As a result, we observe [variance](https://censoredplanet.org/assets/russia.pdf) in how internet censorship is implemented across Russia, as blocks are not implemented on all networks in the country, and different ISPs adopt different censorship techniques. +Russia implements internet censorship in a [decentralized](https://www.ndss-symposium.org/ndss-paper/decentralized-control-a-case-study-of-russia/) manner, as every Internet Service Provider (ISP) in Russia is responsible for implementing government-mandated blocks independently. As a result, we observe [variance](https://censoredplanet.org/assets/russia.pdf) in how internet censorship is implemented across Russia, as blocks are not implemented on all networks in the country, and different ISPs adopt different censorship techniques. To block websites, OONI data shows that Russian ISPs adopt the following censorship techniques: @@ -84,51 +84,51 @@ To block websites, OONI data shows that Russian ISPs adopt the following censors The most common method we see implemented (in terms of the number of ISPs relying on it) is the injection of a RST packet following the ClientHello during the TLS handshake. DNS based filtering is also very prevalent. -Not only does internet censorship vary across providers in Russia, but OONI data also suggests that censorship can also vary in how it’s implemented within an ASN. On some networks where we see block pages being served for a tested website (thereby confirming the block), we see other measurements collected from the same network (on the same day) that show that the same site is accessible. This suggests that on some networks, the blocking of websites is not implemented consistently on an ASN-level, but may vary in its implementation within the same network. Moreover, certain ISPs have implemented blocking through the use of multiple techniques at the same time, making circumvention harder. +Not only does internet censorship vary across providers in Russia, but OONI data also suggests that censorship can also vary in how it’s implemented within an ASN. On some networks where we see block pages being served for a tested website (thereby confirming the block), we see other measurements collected from the same network (on the same day) that show that the same site is accessible. This suggests that on some networks, the blocking of websites is not implemented consistently on an ASN-level, but may vary in its implementation within the same network. Moreover, certain ISPs have implemented blocking through the use of multiple techniques at the same time, making circumvention harder. -But Russia is not the only country to have increased its censorship amid the war in Ukraine. Following [EU sanctions on Russian state-owned media outlets](https://www.consilium.europa.eu/en/press/press-releases/2022/03/02/eu-imposes-sanctions-on-state-owned-outlets-rt-russia-today-and-sputnik-s-broadcasting-in-the-eu/), several European countries (such as [Greece](https://explorer.ooni.org/measurement/20220304T144806Z_webconnectivity_GR_3329_n1_dsFVoQ4VJpoCt8fI?input=https%3A%2F%2Fwww.rt.com) and [France](https://explorer.ooni.org/measurement/20220306T025735Z_webconnectivity_FR_15557_n1_KvMNzvez6OP3omZT?input=https%3A%2F%2Fwww.rt.com%2F)) have recently started blocking access to the Russia Today (RT) and Sputnik News websites. Both Russia and the EU argue that these new censorship events are on the grounds of limiting the spread of misinformation. +But Russia is not the only country to have increased its censorship amid the war in Ukraine. Following [EU sanctions on Russian state-owned media outlets](https://www.consilium.europa.eu/en/press/press-releases/2022/03/02/eu-imposes-sanctions-on-state-owned-outlets-rt-russia-today-and-sputnik-s-broadcasting-in-the-eu/), several European countries (such as [Greece](https://explorer.ooni.org/measurement/20220304T144806Z_webconnectivity_GR_3329_n1_dsFVoQ4VJpoCt8fI?input=https%3A%2F%2Fwww.rt.com) and [France](https://explorer.ooni.org/measurement/20220306T025735Z_webconnectivity_FR_15557_n1_KvMNzvez6OP3omZT?input=https%3A%2F%2Fwww.rt.com%2F)) have recently started blocking access to the Russia Today (RT) and Sputnik News websites. Both Russia and the EU argue that these new censorship events are on the grounds of limiting the spread of misinformation. ### Russia #### Blocking of news media websites -On 24th February 2022, Roskomnadzor published a [statement](https://rkn.gov.ru/news/rsoc/news74084.htm) claiming that media organizations are obliged to “only use information and data received from official Russian sources”. In the same statement, Roskomnadzor reminds the media that “the dissemination of knowingly false information entails liability under Article 13.15 of the Code of Administrative Offenses of the Russian Federation in the form of an administrative fine of up to 5 million rubles”. They also warned that the dissemination of false information on the Internet will result in the immediate blocking of such materials by Roskomnadzor in accordance with Article 15.3 of the Federal Law No. 149-FZ ("On Information, Information Technologies and Information Protection"). +On 24th February 2022, Roskomnadzor published a [statement](https://rkn.gov.ru/news/rsoc/news74084.htm) claiming that media organizations are obliged to “only use information and data received from official Russian sources”. In the same statement, Roskomnadzor reminds the media that “the dissemination of knowingly false information entails liability under Article 13.15 of the Code of Administrative Offenses of the Russian Federation in the form of an administrative fine of up to 5 million rubles”. They also warned that the dissemination of false information on the Internet will result in the immediate blocking of such materials by Roskomnadzor in accordance with Article 15.3 of the Federal Law No. 149-FZ ("On Information, Information Technologies and Information Protection"). -On 4th March 2022, Roskomnadzor [reportedly began blocking access to several foreign news media websites](https://cpj.org/2022/03/russia-further-blocks-media-outlets-social-media/) (including [BBC](https://www.theguardian.com/media/2022/mar/04/bbc-website-blocked-in-russia-as-shortwave-radio-brought-back-to-cover-ukraine-war), [Deutsche Welle](https://www.dw.com/en/dw-website-blocked-in-russia-letter-from-director-general-to-users/a-61011476), [Voice of America](https://www.voanews.com/a/voa-bbc-vow-to-keep-news-flowing-despite-russian-ban-/6470590.html), [Meduza](https://meduza.io/en/feature/2022/03/04/we-ain-t-done-yet), and several services of [Radio Free Europe/Radio Liberty](https://www.rferl.org/a/31735597.html)) on the grounds of spreading false information. On the same day, Russia [adopted a bill with amendments to the Russian Criminal Code](https://interfax.com/newsroom/top-stories/75419/) which introduces higher penalties – including a [prison sentence of up to 15 years](https://cpj.org/2022/03/proposed-russian-legislation-threatens-15-years-in-prison-for-fake-information-about-ukraine-invasion/) –  for those convicted of disseminating false information about Russian military operations, discrediting the Russian Armed Forces, and calls for anti-Russian sanctions. This has resulted in western media outlets [suspending their reporting and broadcasts](https://www.washingtonpost.com/media/2022/03/05/cnn-russia-abc-putin-censorship-law/) from within Russia in order to protect their journalists. Tim Davie, the BBC’s Director General, reportedly [stated](https://www.theguardian.com/media/2022/mar/04/bbc-temporarily-suspending-work-all-news-journalists-russia) the legislation “appears to criminalize the process of independent journalism” in Russia. +On 4th March 2022, Roskomnadzor [reportedly began blocking access to several foreign news media websites](https://cpj.org/2022/03/russia-further-blocks-media-outlets-social-media/) (including [BBC](https://www.theguardian.com/media/2022/mar/04/bbc-website-blocked-in-russia-as-shortwave-radio-brought-back-to-cover-ukraine-war), [Deutsche Welle](https://www.dw.com/en/dw-website-blocked-in-russia-letter-from-director-general-to-users/a-61011476), [Voice of America](https://www.voanews.com/a/voa-bbc-vow-to-keep-news-flowing-despite-russian-ban-/6470590.html), [Meduza](https://meduza.io/en/feature/2022/03/04/we-ain-t-done-yet), and several services of [Radio Free Europe/Radio Liberty](https://www.rferl.org/a/31735597.html)) on the grounds of spreading false information. On the same day, Russia [adopted a bill with amendments to the Russian Criminal Code](https://interfax.com/newsroom/top-stories/75419/) which introduces higher penalties – including a [prison sentence of up to 15 years](https://cpj.org/2022/03/proposed-russian-legislation-threatens-15-years-in-prison-for-fake-information-about-ukraine-invasion/) – for those convicted of disseminating false information about Russian military operations, discrediting the Russian Armed Forces, and calls for anti-Russian sanctions. This has resulted in western media outlets [suspending their reporting and broadcasts](https://www.washingtonpost.com/media/2022/03/05/cnn-russia-abc-putin-censorship-law/) from within Russia in order to protect their journalists. Tim Davie, the BBC’s Director General, reportedly [stated](https://www.theguardian.com/media/2022/mar/04/bbc-temporarily-suspending-work-all-news-journalists-russia) the legislation “appears to criminalize the process of independent journalism” in Russia. We are able to observe the blocking of media websites in Russia through OONI data, as illustrated in the chart below. ![](images/image20.png) -Chart: Blocking of news media websites in Russia based on OONI measurements (collected between February to March 2022), [https://explorer.ooni.org/experimental/mat?probe_cc=RU&test_name=web_connectivity&category_code=NEWS&since=2022-02-02&until=2022-03-05&axis_x=measurement_start_day&axis_y=input](https://explorer.ooni.org/experimental/mat?probe_cc=RU&test_name=web_connectivity&category_code=NEWS&since=2022-02-02&until=2022-03-05&axis_x=measurement_start_day&axis_y=input)  +**Chart:** Blocking of news media websites in Russia based on OONI measurements (collected between February to March 2022), [https://explorer.ooni.org/experimental/mat?probe_cc=RU&test_name=web_connectivity&category_code=NEWS&since=2022-02-02&until=2022-03-05&axis_x=measurement_start_day&axis_y=input](https://explorer.ooni.org/experimental/mat?probe_cc=RU&test_name=web_connectivity&category_code=NEWS&since=2022-02-02&until=2022-03-05&axis_x=measurement_start_day&axis_y=input) The above chart aggregates OONI measurements collected from Russia (between February 2022 to March 2022) pertaining to the testing of news media websites that started presenting anomalies following the invasion of Ukraine (which started on 24th February 2022). Each bar for a given domain and day represents the percentage of measurements for a given status (`anomaly`, `confirmed` or `OK`) aggregated across all networks. As is evident from the above chart, these media websites were found accessible (i.e. green) on almost all tested networks in Russia over the last month, and only started presenting signs of blocking (“anomalies”, indicated in orange) following the invasion of Ukraine. It is also evident that these media websites are not blocked on all tested networks in Russia, as we observe that they were found accessible on some networks. -Moreover, our analysis (through the above chart) shows that the blocking of news media websites started on different days. First we observe the [blocking of Current Time TV](https://explorer.ooni.org/search?since=2022-02-27&until=2022-02-28&failure=false&domain=www.currenttime.tv&probe_cc=RU&only=anomalies) (currentime.tv), [Crimea Realities](https://explorer.ooni.org/search?since=2022-02-27&until=2022-02-28&failure=false&domain=ru.krymr.com&probe_cc=RU&only=anomalies) (ru.krymr.com) and [Interfax](https://explorer.ooni.org/search?since=2022-02-27&until=2022-02-28&failure=false&domain=www.interfax.ru&probe_cc=RU&only=anomalies) (interfax.ru), which appears to have started on 27th February 2022. The timing of the block (as suggested by OONI data) is also corroborated by Radio Free Europe/Radio Liberty (RFE/RL), who published a [statement](https://pressroom.rferl.org/a/rferl-strongly-condemns-blockage-of-russian-language-websites-and-harassment-of-journalists/31728748.html) about the blocking of currentime.tv and ru.krymr.com on 28th February 2022. We then observe that the blocking of tvrain.ru [began](https://explorer.ooni.org/search?since=2022-03-01&until=2022-03-02&failure=false&domain=tvrain.ru&probe_cc=RU&only=anomalies) on 1st March 2022. The blocking of all other media websites listed in the above graph appears to have [started](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-05&failure=false&probe_cc=RU&only=anomalies&test_name=web_connectivity) on the night of 3rd-4th March 2022. +Moreover, our analysis (through the above chart) shows that the blocking of news media websites started on different days. First we observe the [blocking of Current Time TV](https://explorer.ooni.org/search?since=2022-02-27&until=2022-02-28&failure=false&domain=www.currenttime.tv&probe_cc=RU&only=anomalies) (currentime.tv), [Crimea Realities](https://explorer.ooni.org/search?since=2022-02-27&until=2022-02-28&failure=false&domain=ru.krymr.com&probe_cc=RU&only=anomalies) (ru.krymr.com) and [Interfax](https://explorer.ooni.org/search?since=2022-02-27&until=2022-02-28&failure=false&domain=www.interfax.ru&probe_cc=RU&only=anomalies) (interfax.ru), which appears to have started on 27th February 2022. The timing of the block (as suggested by OONI data) is also corroborated by Radio Free Europe/Radio Liberty (RFE/RL), who published a [statement](https://pressroom.rferl.org/a/rferl-strongly-condemns-blockage-of-russian-language-websites-and-harassment-of-journalists/31728748.html) about the blocking of currentime.tv and ru.krymr.com on 28th February 2022. We then observe that the blocking of tvrain.ru [began](https://explorer.ooni.org/search?since=2022-03-01&until=2022-03-02&failure=false&domain=tvrain.ru&probe_cc=RU&only=anomalies) on 1st March 2022. The blocking of all other media websites listed in the above graph appears to have [started](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-05&failure=false&probe_cc=RU&only=anomalies&test_name=web_connectivity) on the night of 3rd-4th March 2022. ##### BBC -On 4th March 2022,  it was [reported](https://www.theguardian.com/media/2022/mar/04/bbc-website-blocked-in-russia-as-shortwave-radio-brought-back-to-cover-ukraine-war) that ISPs in Russia began blocking access to BBC, with their website reportedly being available at only 17% of normal levels in Russia. +On 4th March 2022, it was [reported](https://www.theguardian.com/media/2022/mar/04/bbc-website-blocked-in-russia-as-shortwave-radio-brought-back-to-cover-ukraine-war) that ISPs in Russia began blocking access to BBC, with their website reportedly being available at only 17% of normal levels in Russia. This is corroborated by OONI data. At 21:34 UTC on 3rd March 2022, OONI measurements collected from Russia started to show the [first signs of BBC blocking](https://explorer.ooni.org/measurement/20220303T213429Z_webconnectivity_RU_41668_n1_krRshjjBSYrOQK0l?input=https%3A%2F%2Fwww.bbc.com%2Fnews). On 4th March 2022, most OONI measurements from the testing of bbc.com on multiple networks in Russia [presented signs of blocking](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&probe_cc=RU&domain=www.bbc.com&only=anomalies), as illustrated below. ![](images/image17.png) -Chart: OONI data on the testing of bbc.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.bbc.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.bbc.com&probe_cc=RU)  +**Chart:** OONI data on the testing of bbc.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.bbc.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.bbc.com&probe_cc=RU) -The OONI Probe testing of bbc.com in Russia [shows](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&probe_cc=RU&domain=www.bbc.com) that different ISPs block access to the site using different censorship techniques, and that bbc.com is still accessible on some networks in Russia. +The OONI Probe testing of bbc.com in Russia [shows](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&probe_cc=RU&domain=www.bbc.com) that different ISPs block access to the site using different censorship techniques, and that bbc.com is still accessible on some networks in Russia. -Based on OONI data, we are able to [automatically confirm the blocking of bbc.com](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&probe_cc=RU&domain=www.bbc.com&only=confirmed) on the following 17 networks in Russia: `AS34533`, `AS31213`, `AS48642`, `AS3216`, `AS13335`, `AS8492`, `AS31133`, `AS31257`, `AS25513`, `AS42668`, `AS8402`, `AS8427`, `AS8359`, `AS50542`, `AS41733`, `AS12389`, `AS52207`. On these networks, OONI data shows that a [block page is served](https://explorer.ooni.org/measurement/20220305T093618Z_webconnectivity_RU_41733_n1_oOXBvcd4vbRZrnPh?input=http%3A%2F%2Fwww.bbc.com%2Fnews) when internet users try to access bbc.com. +Based on OONI data, we are able to [automatically confirm the blocking of bbc.com](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&probe_cc=RU&domain=www.bbc.com&only=confirmed) on the following 17 networks in Russia: `AS34533`, `AS31213`, `AS48642`, `AS3216`, `AS13335`, `AS8492`, `AS31133`, `AS31257`, `AS25513`, `AS42668`, `AS8402`, `AS8427`, `AS8359`, `AS50542`, `AS41733`, `AS12389`, `AS52207`. On these networks, OONI data shows that a [block page is served](https://explorer.ooni.org/measurement/20220305T093618Z_webconnectivity_RU_41733_n1_oOXBvcd4vbRZrnPh?input=http%3A%2F%2Fwww.bbc.com%2Fnews) when internet users try to access bbc.com. ![](images/image11.png) -Image: Block page served for bbc.com on AS41733 network in Russia. +**Image:** Block page served for bbc.com on AS41733 network in Russia. We observe DNS based interference on 14 different networks, as illustrated by the following table: -|probe_asn|blocking_method        |count| +|probe_asn|blocking_method|count| |---------|-----------------------|-----| |AS12389  |dns.confirmed          |1    | |         |http.confirmed_pipeline|5    | @@ -158,7 +158,7 @@ We observe DNS based interference on 14 different networks, as illustrated by th |         |https.generic_failure  |3    | |AS8790   |dns.confirmed          |16   | -On some networks, such as AS12389, we notice that the blocking is implemented using three distinct methods. For example, if we look at an [OONI measurement collected on 4th March 2022](https://explorer.ooni.org/measurement/20220304T044651Z_webconnectivity_RU_12389_n1_BmRicVwEpc4HG72k?input=https%3A%2F%2Fwww.bbc.com%2F), we can see that the DNS query produces an inconsistent result (the IP `31.28.24.3` which is known to serve a block page), yet when the probe performs a TLS handshake with `31.28.24.3` using the SNI for [www.bbc.com](http://www.bbc.com) we see a RST packet being injected. We exclude that this behavior is of the destination host when an unrecognized SNI is presented as we are able to perform a successful TLS handshake from a non-local vantage point in Italy: +On some networks, such as AS12389, we notice that the blocking is implemented using three distinct methods. For example, if we look at an [OONI measurement collected on 4th March 2022](https://explorer.ooni.org/measurement/20220304T044651Z_webconnectivity_RU_12389_n1_BmRicVwEpc4HG72k?input=https%3A%2F%2Fwww.bbc.com%2F), we can see that the DNS query produces an inconsistent result (the IP `31.28.24.3` which is known to serve a block page), yet when the probe performs a TLS handshake with `31.28.24.3` using the SNI for [www.bbc.com](http://www.bbc.com) we see a RST packet being injected. We exclude that this behavior is of the destination host when an unrecognized SNI is presented as we are able to perform a successful TLS handshake from a non-local vantage point in Italy: ``` % openssl s_client -connect 31.28.24.3:443 -servername www.bbc.com @@ -180,7 +180,7 @@ This leads us to conclude that multiple blocking technologies are being employed On 17 networks in Russia we see that a block page known by our data processing pipeline is served when performing an HTTP request: -|probe_asn|blocking_method        |count| +|probe_asn|blocking_method|count| |---------|-----------------------|-----| |AS12389  |dns.confirmed          |1    | |         |http.confirmed_pipeline|5    | @@ -238,7 +238,7 @@ Also in this case, we can see on many networks the presence of `tls.connection_r On 6 networks in Russia, we notice the presence of a TLS man-in-the-middle: -|probe_asn|blocking_method        |count| +|probe_asn|blocking_method|count| |---------|-----------------------|-----| |AS202422 |tls.mitm               |1    | |AS25513  |http.confirmed_pipeline|2    | @@ -296,7 +296,7 @@ We then look at the breakdown of measurements that show the connection being res This blocking pattern is seen on a total of 30 different networks, as illustrated in the table below: -|probe_asn|blocking_method        |count| +|probe_asn|blocking_method|count| |---------|-----------------------|-----| |AS12389  |http.confirmed_pipeline|5    | |         |https.connection_reset |10   | @@ -340,7 +340,7 @@ This blocking pattern is seen on a total of 30 different networks, as illustrate |AS8492   |tls.timeout |5    | |AS8732   |tls.connection_reset   |3    | -Based on these results, it appears that this blocking technique is the most prevalent one. The determination of the `tls.connection_reset` blocking method was produced by analyzing the `network_events` keys of [OONI Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity) and looking for a `connection_reset` failure event following the first write operation during the TLS handshake, which is the ClientHello containing the SNI field. +Based on these results, it appears that this blocking technique is the most prevalent one. The determination of the `tls.connection_reset` blocking method was produced by analyzing the `network_events` keys of [OONI Web Connectivity measurements](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&probe_cc=RU&test_name=web_connectivity) and looking for a `connection_reset` failure event following the first write operation during the TLS handshake, which is the ClientHello containing the SNI field. This is a strong indication that the blocking is happening on an SNI basis. @@ -377,7 +377,7 @@ On some networks we observe, though, that the filtering seems to not only be bas The connection is reset when performing a handshake towards `151.101.192.81` and `151.101.64.81`, but not when the same operation is done on `151.101.0.81` and `151.101.128.81`. -According to [zapret-info](https://github.com/zapret-info/z-i) (a public dump of the Russian blocklist) all of the above IPs were added to the blocklist for [www.bbc.com](http://www.bbc.com): +According to [zapret-info](https://github.com/zapret-info/z-i) (a public dump of the Russian blocklist) all of the above IPs were added to the blocklist for [www.bbc.com](http://www.bbc.com): ``` 8630:151.101.0.81 | 151.101.128.81 | 151.101.192.81 | 151.101.64.81;\*.www.bbc.com;; @@ -422,19 +422,19 @@ This is seen on 4 different networks: ##### Deutsche Welle -In early February 2022, [Russia banned Deutsche Welle (DW) from broadcasting in Russia](https://www.dw.com/en/dw-ban-in-russia-how-it-happened/a-60666435), in response to Germany blocking Russian state broadcaster RT for not having a broadcasting license. However, Moscow’s reaction to ban DW was viewed as disproportionate, given that DW correspondents lost their accreditation in Russia even though DW has an official broadcast license in Russia for its channels, while RT journalists were still allowed to work in Germany. +In early February 2022, [Russia banned Deutsche Welle (DW) from broadcasting in Russia](https://www.dw.com/en/dw-ban-in-russia-how-it-happened/a-60666435), in response to Germany blocking Russian state broadcaster RT for not having a broadcasting license. However, Moscow’s reaction to ban DW was viewed as disproportionate, given that DW correspondents lost their accreditation in Russia even though DW has an official broadcast license in Russia for its channels, while RT journalists were still allowed to work in Germany. -Similarly to BBC, access to Deutsche Welle’s website (www.dw.com) was reportedly blocked in Russia on 4th March 2022. Peter Limbourg, the Director General of Deutsche Welle, [published a letter](https://www.dw.com/en/dw-website-blocked-in-russia-letter-from-director-general-to-users/a-61011476) stating that their website had been blocked in all broadcast languages in Russia, asking DW readers to [use censorship circumvention tools](https://www.dw.com/en/tor-psiphon-signal-and-co-how-to-move-unrecognized-on-the-internet/a-51720615) to access DW channels. DW’s website can also be accessed through the use of a [Tor onion address](https://www.dw.com/en/dw-websites-accessible-via-tor-protocol/a-51338328). +Similarly to BBC, access to Deutsche Welle’s website (www.dw.com) was reportedly blocked in Russia on 4th March 2022. Peter Limbourg, the Director General of Deutsche Welle, [published a letter](https://www.dw.com/en/dw-website-blocked-in-russia-letter-from-director-general-to-users/a-61011476) stating that their website had been blocked in all broadcast languages in Russia, asking DW readers to [use censorship circumvention tools](https://www.dw.com/en/tor-psiphon-signal-and-co-how-to-move-unrecognized-on-the-internet/a-51720615) to access DW channels. DW’s website can also be accessed through the use of a [Tor onion address](https://www.dw.com/en/dw-websites-accessible-via-tor-protocol/a-51338328). -The blocking of dw.com is corroborated by OONI data, which shows that while dw.com was previously accessible in Russia, it [started showing signs of blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-04&failure=false&probe_cc=RU&domain=www.dw.com) on 4th March 2022, as illustrated below. +The blocking of dw.com is corroborated by OONI data, which shows that while dw.com was previously accessible in Russia, it [started showing signs of blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-04&failure=false&probe_cc=RU&domain=www.dw.com) on 4th March 2022, as illustrated below. ![](images/image1.png) -Chart: OONI data on the testing of dw.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.dw.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.dw.com&probe_cc=RU) +**Chart:** OONI data on the testing of dw.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.dw.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.dw.com&probe_cc=RU) The above graph aggregates OONI measurement coverage (pertaining to the testing of dw.com) from multiple AS networks in Russia between early February 2022 to March 2022. The sudden and large volume of anomalies on 4th March 2022 provides a signal of interference, though it’s worth highlighting that dw.com still appeared to be accessible on some networks in Russia. -Based on our analysis of anomalous measurements, we provide a table below that summarizes the ways through which access to dw.com is being [blocked](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&failure=false&probe_cc=RU&domain=www.dw.com&only=anomalies) in Russia. +Based on our analysis of anomalous measurements, we provide a table below that summarizes the ways through which access to dw.com is being [blocked](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&failure=false&probe_cc=RU&domain=www.dw.com&only=anomalies) in Russia. |probe_asn|blocking_recalc       |count| |---------|----------------------|-----| @@ -517,7 +517,7 @@ Based on our analysis of anomalous measurements, we provide a table below that s |AS31163  |tls.connection_closed |13   | |         |tls.timeout|1    | -Specifically, we see a **strong indication of blocking on at least 54 distinct ASNs** and we are able to confirm the blocking (though DNS based fingerprints) on 16 networks and see a TLS man-in-the-middle on 4 of them. +Specifically, we see a **strong indication of blocking on at least 54 distinct ASNs** and we are able to confirm the blocking (though DNS based fingerprints) on 16 networks and see a TLS man-in-the-middle on 4 of them. As the patterns and specifics of the blocking techniques are similar to those seen for bbc.com, we recommend referring to that section for a more in-depth analysis. @@ -533,81 +533,81 @@ Below we share OONI data on the blocking of websites run by RFE/RL services. On 4th March 2022, the Russian service of Radio Liberty [reported](https://www.svoboda.org/a/sayt-radio-svoboda-zablokirovan-v-rossii/31732804.html) that access to their website (svoboda.org) was blocked in Russia (which was confirmed by Roskomnadzor, along with the blocking of other media websites discussed in this report). -OONI data corroborates both the [blocking of svoboda.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.svoboda.org&probe_cc=RU) in Russia and the timing of the block. +OONI data corroborates both the [blocking of svoboda.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.svoboda.org&probe_cc=RU) in Russia and the timing of the block. ![](images/image22.png) -Chart: OONI data on the testing of svoboda.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.svoboda.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.svoboda.org&probe_cc=RU)  +**Chart:** OONI data on the testing of svoboda.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.svoboda.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.svoboda.org&probe_cc=RU) -The above graph aggregates OONI measurement coverage from the testing of svoboda.org on multiple ASNs in Russia between early February 2022 to early March 2022. It is evident that the testing of svoboda.org only [started to present signs of blocking](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-05&failure=false&domain=www.svoboda.org&probe_cc=RU) in the early hours of 4th March 2022, while most other measurements collected previously show that svoboda.org used to be accessible on many networks in Russia. It’s worth highlighting that the [blocking of svoboda.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.svoboda.org&probe_cc=RU) varies across ISPs in Russia, and we observe many different censorship techniques (such as [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220306T134733Z_webconnectivity_RU_31163_n1_srGrkRq3HprfJLoS?input=https%3A%2F%2Fwww.svoboda.org%2F) and [connection resets](https://explorer.ooni.org/measurement/20220306T135204Z_webconnectivity_RU_42387_n1_WMsbYmBIDAWpawBM?input=https%3A%2F%2Fwww.svoboda.org%2F)). The site still appears to be [accessible](https://explorer.ooni.org/measurement/20220306T135016Z_webconnectivity_RU_34757_n1_DNlTue65wo3iJLqO?input=https%3A%2F%2Fwww.svoboda.org%2F) on some networks in Russia. +The above graph aggregates OONI measurement coverage from the testing of svoboda.org on multiple ASNs in Russia between early February 2022 to early March 2022. It is evident that the testing of svoboda.org only [started to present signs of blocking](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-05&failure=false&domain=www.svoboda.org&probe_cc=RU) in the early hours of 4th March 2022, while most other measurements collected previously show that svoboda.org used to be accessible on many networks in Russia. It’s worth highlighting that the [blocking of svoboda.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.svoboda.org&probe_cc=RU) varies across ISPs in Russia, and we observe many different censorship techniques (such as [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220306T134733Z_webconnectivity_RU_31163_n1_srGrkRq3HprfJLoS?input=https%3A%2F%2Fwww.svoboda.org%2F) and [connection resets](https://explorer.ooni.org/measurement/20220306T135204Z_webconnectivity_RU_42387_n1_WMsbYmBIDAWpawBM?input=https%3A%2F%2Fwww.svoboda.org%2F)). The site still appears to be [accessible](https://explorer.ooni.org/measurement/20220306T135016Z_webconnectivity_RU_34757_n1_DNlTue65wo3iJLqO?input=https%3A%2F%2Fwww.svoboda.org%2F) on some networks in Russia. **Krym.Realii** -[Krym.Realii](https://ru.krymr.com/) is the Crimean project of the Ukrainian service of Radio Liberty that was launched following the annexation of the Crimean Peninsula by Russia. +[Krym.Realii](https://ru.krymr.com/) is the Crimean project of the Ukrainian service of Radio Liberty that was launched following the annexation of the Crimean Peninsula by Russia. -On 28th February 2022, Radio Liberty [reported](https://pressroom.rferl.org/a/rferl-strongly-condemns-blockage-of-russian-language-websites-and-harassment-of-journalists/31728748.html) that Roskomnadzor blocked access to the Russian-language site of Krym.Realii (`ru.krymr.com`). This is corroborated by OONI data, which [shows](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&probe_cc=RU&domain=ru.krymr.com&only=anomalies) that the testing of `ru.krymr.com` presented an increased volume of anomalies from 27th February 2022 onwards (in comparison to previous months), suggesting that access to the site was blocked on some networks in Russia. +On 28th February 2022, Radio Liberty [reported](https://pressroom.rferl.org/a/rferl-strongly-condemns-blockage-of-russian-language-websites-and-harassment-of-journalists/31728748.html) that Roskomnadzor blocked access to the Russian-language site of Krym.Realii (`ru.krymr.com`). This is corroborated by OONI data, which [shows](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&probe_cc=RU&domain=ru.krymr.com&only=anomalies) that the testing of `ru.krymr.com` presented an increased volume of anomalies from 27th February 2022 onwards (in comparison to previous months), suggesting that access to the site was blocked on some networks in Russia. ![](images/image9.png) -Chart: OONI data on the testing of ru.krymr.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=ru.krymr.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=ru.krymr.com&probe_cc=RU)  +**Chart:** OONI data on the testing of ru.krymr.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=ru.krymr.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=ru.krymr.com&probe_cc=RU)  -The blocking of `ru.krymr.com` is [automatically confirmed](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=ru.krymr.com&probe_cc=RU&only=confirmed) on AS51813 in Russia, where OONI data [shows](https://explorer.ooni.org/measurement/20220305T163057Z_webconnectivity_RU_51813_n1_IMd0vUCbilW3jp95?input=https%3A%2F%2Fru.krymr.com%2F) that DNS resolution leads to the IP `62.33.207.196` which hosts the following Russian block page: +The blocking of `ru.krymr.com` is [automatically confirmed](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=ru.krymr.com&probe_cc=RU&only=confirmed) on AS51813 in Russia, where OONI data [shows](https://explorer.ooni.org/measurement/20220305T163057Z_webconnectivity_RU_51813_n1_IMd0vUCbilW3jp95?input=https%3A%2F%2Fru.krymr.com%2F) that DNS resolution leads to the IP `62.33.207.196` which hosts the following Russian block page: ![](images/image21.png) -Image: Block page served for ru.krymr.com on AS51813 in Russia. +**Image:** Block page served for ru.krymr.com on AS51813 in Russia. -On other networks, OONI data shows that access to `ru.krymr.com` is blocked by a variety of censorship techniques, such as [connections being reset](https://explorer.ooni.org/measurement/20220305T180355Z_webconnectivity_RU_52207_n1_92jLVjn2CcESOUDc?input=https%3A%2F%2Fru.krymr.com%2F) and [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220305T164647Z_webconnectivity_RU_31163_n1_GlL89tCA0gIDZjgX?input=https%3A%2F%2Fru.krymr.com%2Fradio%2Fschedule%2F103%2F). Krym.Realii [published](https://ru.krymr.com/a/news-krym-realii-blokirovka-roskomnadzor/31727577.html) a blog post on the blocking of their website, sharing their [mirror site](https://rukrymr.azureedge.net/) for censorship circumvention. +On other networks, OONI data shows that access to `ru.krymr.com` is blocked by a variety of censorship techniques, such as [connections being reset](https://explorer.ooni.org/measurement/20220305T180355Z_webconnectivity_RU_52207_n1_92jLVjn2CcESOUDc?input=https%3A%2F%2Fru.krymr.com%2F) and [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220305T164647Z_webconnectivity_RU_31163_n1_GlL89tCA0gIDZjgX?input=https%3A%2F%2Fru.krymr.com%2Fradio%2Fschedule%2F103%2F). Krym.Realii [published](https://ru.krymr.com/a/news-krym-realii-blokirovka-roskomnadzor/31727577.html) a blog post on the blocking of their website, sharing their [mirror site](https://rukrymr.azureedge.net/) for censorship circumvention. **Idel.Realii** -[Idel.Realii](https://www.idelreal.org/) is a media project of the Tatar-Bashkir service of Radio Liberty about the Volga region. +[Idel.Realii](https://www.idelreal.org/) is a media project of the Tatar-Bashkir service of Radio Liberty about the Volga region. -Starting from the early hours of 4th March 2022, OONI data [shows](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-05&failure=false&domain=www.idelreal.org&probe_cc=RU) that multiple ISPs in Russia started blocking access to idelreal.org, as illustrated below. +Starting from the early hours of 4th March 2022, OONI data [shows](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-05&failure=false&domain=www.idelreal.org&probe_cc=RU) that multiple ISPs in Russia started blocking access to idelreal.org, as illustrated below. ![](images/image30.png) -Chart: OONI data on the testing of idelreal.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.idelreal.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.idelreal.org&probe_cc=RU)  +**Chart:** OONI data on the testing of idelreal.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.idelreal.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.idelreal.org&probe_cc=RU) -The above graph aggregates OONI measurement coverage from the testing of idelreal.org on multiple ASNs in Russia between early February 2022 to early March 2022. It is evident that while idelreal.org was previously accessible in Russia, it started presenting signs of [blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&failure=false&domain=www.idelreal.org&probe_cc=RU&only=anomalies) on many networks from the early hours of 4th March 2022. It’s worth noting though that idelreal.org was still [accessible](https://explorer.ooni.org/measurement/20220306T120304Z_webconnectivity_RU_203337_n1_ZHNDGamWG9otw6kv?input=https%3A%2F%2Fwww.idelreal.org%2F) on some networks in Russia. +The above graph aggregates OONI measurement coverage from the testing of idelreal.org on multiple ASNs in Russia between early February 2022 to early March 2022. It is evident that while idelreal.org was previously accessible in Russia, it started presenting signs of [blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&failure=false&domain=www.idelreal.org&probe_cc=RU&only=anomalies) on many networks from the early hours of 4th March 2022. It’s worth noting though that idelreal.org was still [accessible](https://explorer.ooni.org/measurement/20220306T120304Z_webconnectivity_RU_203337_n1_ZHNDGamWG9otw6kv?input=https%3A%2F%2Fwww.idelreal.org%2F) on some networks in Russia. -Idel.Realii provide [information on how to circumvent the blocking](https://www.idelreal.org/p/8513.html) of their website. +Idel.Realii provide [information on how to circumvent the blocking](https://www.idelreal.org/p/8513.html) of their website. **Kavkaz.Realii** -[Kavkaz.Realii](https://www.kavkazr.com/) is a media project of the North Caucasus Service of Radio Free Europe/Radio Liberty. On 4th March 2022, they [reported](https://www.kavkazr.com/a/sayt-kavkaz-realii-zablokirovan-v-rossii-chto-delatj/31734935.html) that access to their website was blocked in Russia, following Roskomnadzor’s demands that their editors remove [news](https://www.kavkazr.com/a/kadyrov-priznal-poteri-sredi-chechenskih-voennyh-v-ukraine/31729658.html) about the losses among military personnel from Chechnya in the war in Ukraine. +[Kavkaz.Realii](https://www.kavkazr.com/) is a media project of the North Caucasus Service of Radio Free Europe/Radio Liberty. On 4th March 2022, they [reported](https://www.kavkazr.com/a/sayt-kavkaz-realii-zablokirovan-v-rossii-chto-delatj/31734935.html) that access to their website was blocked in Russia, following Roskomnadzor’s demands that their editors remove [news](https://www.kavkazr.com/a/kadyrov-priznal-poteri-sredi-chechenskih-voennyh-v-ukraine/31729658.html) about the losses among military personnel from Chechnya in the war in Ukraine. -OONI data corroborates both the [blocking of kavkazr.com](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&input=https%3A%2F%2Fwww.kavkazr.com%2F&domain=www.kavkazr.com&failure=false&only=anomalies) and the timing of the block. +OONI data corroborates both the [blocking of kavkazr.com](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&input=https%3A%2F%2Fwww.kavkazr.com%2F&domain=www.kavkazr.com&failure=false&only=anomalies) and the timing of the block. ![](images/image16.png) -Chart: OONI data on the testing of kavkazr.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.kavkazr.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.kavkazr.com&probe_cc=RU)  +**Chart:** OONI data on the testing of kavkazr.com in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.kavkazr.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.kavkazr.com&probe_cc=RU) -The above graph aggregates OONI measurement coverage from the testing of kavkazr.com on multiple ASNs in Russia between early February 2022 to early March 2022. We can clearly see that while kavkazr.com was previously accessible in Russia, it [started presenting signs of blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-05&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&input=https%3A%2F%2Fwww.kavkazr.com%2F&domain=www.kavkazr.com&failure=false) in the early hours of 4th March 2022 (which has persisted since). Similarly to the blocking of other websites, we see that the [blocking of kavkazr.com](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&input=https%3A%2F%2Fwww.kavkazr.com%2F&domain=www.kavkazr.com&failure=false&only=anomalies) varies from ISP to ISP in Russia (while some ISPs do not block access), and that different censorship techniques are employed (such as [connection resets](https://explorer.ooni.org/measurement/20220306T100850Z_webconnectivity_RU_25513_n1_y0bso0XKm1jpHe2z?input=https%3A%2F%2Fwww.kavkazr.com%2F)).  +The above graph aggregates OONI measurement coverage from the testing of kavkazr.com on multiple ASNs in Russia between early February 2022 to early March 2022. We can clearly see that while kavkazr.com was previously accessible in Russia, it [started presenting signs of blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-05&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&input=https%3A%2F%2Fwww.kavkazr.com%2F&domain=www.kavkazr.com&failure=false) in the early hours of 4th March 2022 (which has persisted since). Similarly to the blocking of other websites, we see that the [blocking of kavkazr.com](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-07&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&input=https%3A%2F%2Fwww.kavkazr.com%2F&domain=www.kavkazr.com&failure=false&only=anomalies) varies from ISP to ISP in Russia (while some ISPs do not block access), and that different censorship techniques are employed (such as [connection resets](https://explorer.ooni.org/measurement/20220306T100850Z_webconnectivity_RU_25513_n1_y0bso0XKm1jpHe2z?input=https%3A%2F%2Fwww.kavkazr.com%2F)). **Sever.Realii** -[Sever.Realii](https://www.severreal.org/) is a media project of the [Russian Service of Radio Liberty](https://www.svoboda.org/contact). Starting from the early hours of 4th March 2022, OONI data starts to show the [blocking of severreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-05&failure=false&domain=www.severreal.org&probe_cc=RU) on multiple networks in Russia, as illustrated below. +[Sever.Realii](https://www.severreal.org/) is a media project of the [Russian Service of Radio Liberty](https://www.svoboda.org/contact). Starting from the early hours of 4th March 2022, OONI data starts to show the [blocking of severreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-05&failure=false&domain=www.severreal.org&probe_cc=RU) on multiple networks in Russia, as illustrated below. ![](images/image27.png) -Chart: OONI data on the testing of severreal.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.severreal.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.severreal.org&probe_cc=RU)  +**Chart:** OONI data on the testing of severreal.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.severreal.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.severreal.org&probe_cc=RU) -The above graph aggregates OONI measurement coverage from the testing of severreal.org on multiple ASNs in Russia between early February 2022 to early March 2022. We can clearly see that the testing of severreal.org only started to present signs of blocking from the early hours of 4th March 2022 onwards. Similarly to the blocking of other websites, we see that the [blocking of severreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.severreal.org&probe_cc=RU) varies from ISP to ISP in Russia (while some ISPs do not block access), and that different censorship techniques are employed (we primarily see [connections being reset](https://explorer.ooni.org/measurement/20220306T102640Z_webconnectivity_RU_42610_n1_90cWPpqSoxbryShI?input=https%3A%2F%2Fwww.severreal.org%2F)). +The above graph aggregates OONI measurement coverage from the testing of severreal.org on multiple ASNs in Russia between early February 2022 to early March 2022. We can clearly see that the testing of severreal.org only started to present signs of blocking from the early hours of 4th March 2022 onwards. Similarly to the blocking of other websites, we see that the [blocking of severreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.severreal.org&probe_cc=RU) varies from ISP to ISP in Russia (while some ISPs do not block access), and that different censorship techniques are employed (we primarily see [connections being reset](https://explorer.ooni.org/measurement/20220306T102640Z_webconnectivity_RU_42610_n1_90cWPpqSoxbryShI?input=https%3A%2F%2Fwww.severreal.org%2F)). -Sever.Realii published [information on how to circumvent the blocking](https://www.severreal.org/block) of their website. +Sever.Realii published [information on how to circumvent the blocking](https://www.severreal.org/block) of their website. **Siberia.Realii** -[Siberia.Realii](https://www.sibreal.org/) is a media project of the [Russian service of Radio Liberty](https://www.svoboda.org/contact). Starting from the early hours of 4th March 2022, OONI data starts to show the [blocking of sibreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-05&failure=false&domain=www.sibreal.org&probe_cc=RU) on multiple networks in Russia, as illustrated below. +[Siberia.Realii](https://www.sibreal.org/) is a media project of the [Russian service of Radio Liberty](https://www.svoboda.org/contact). Starting from the early hours of 4th March 2022, OONI data starts to show the [blocking of sibreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-05&failure=false&domain=www.sibreal.org&probe_cc=RU) on multiple networks in Russia, as illustrated below. ![](images/image28.png) -Chart: OONI data on the testing of sibreal.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.sibreal.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.sibreal.org&probe_cc=RU)  +**Chart:** OONI data on the testing of sibreal.org in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.sibreal.org&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.sibreal.org&probe_cc=RU) -The above graph aggregates OONI measurement coverage from the testing of sibreal.org on multiple ASNs in Russia between early February 2022 to early March 2022. We can clearly see that the testing of sibreal.org only started to present signs of blocking from the early hours of 4th March 2022 onwards. Similarly to the blocking of other websites, we see that the [blocking of sibreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.sibreal.org&probe_cc=RU) varies from ISP to ISP in Russia (while some ISPs do not block access), and that different censorship techniques are employed (we primarily see [connections being reset](https://explorer.ooni.org/measurement/20220306T100850Z_webconnectivity_RU_25513_n1_y0bso0XKm1jpHe2z?input=https%3A%2F%2Fwww.sibreal.org%2F)). +The above graph aggregates OONI measurement coverage from the testing of sibreal.org on multiple ASNs in Russia between early February 2022 to early March 2022. We can clearly see that the testing of sibreal.org only started to present signs of blocking from the early hours of 4th March 2022 onwards. Similarly to the blocking of other websites, we see that the [blocking of sibreal.org](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.sibreal.org&probe_cc=RU) varies from ISP to ISP in Russia (while some ISPs do not block access), and that different censorship techniques are employed (we primarily see [connections being reset](https://explorer.ooni.org/measurement/20220306T100850Z_webconnectivity_RU_25513_n1_y0bso0XKm1jpHe2z?input=https%3A%2F%2Fwww.sibreal.org%2F)). -Siberia.Realii published [information on how to circumvent the blocking](https://www.sibreal.org/block) of their website. +Siberia.Realii published [information on how to circumvent the blocking](https://www.sibreal.org/block) of their website. ##### Voice of America @@ -615,89 +615,89 @@ On 2nd March 2022, Voice of America (VOA) [reported](https://www.voanews.com/a/r ![](images/image15.png) -Chart: OONI data on the testing of VOA in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU)  +**Chart:** OONI data on the testing of VOA in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU) -The above graph aggregates OONI measurement coverage from the testing of golos-ameriki.ru on multiple ASNs in Russia between early February 2022 to early March 2022. While OONI data shows that the site was previously accessible in Russia, we start to observe [blocking of golos-ameriki.ru](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU) from 4th March 2022 onwards. On many networks, we observe the [injection of a RST packet](https://explorer.ooni.org/measurement/20220306T142058Z_webconnectivity_RU_44943_n1_1BuqOmMQy0KMhnp7?input=https%3A%2F%2Fwww.golos-ameriki.ru%2F) after the ClientHello during the TLS handshake, while on other networks, we observe a [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220306T034232Z_webconnectivity_RU_31163_n1_8ng4wgrawcikw9qp?input=https%3A%2F%2Fwww.golos-ameriki.ru%2F). The block [varies](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU) across networks, and the site remains accessible on some networks. +The above graph aggregates OONI measurement coverage from the testing of golos-ameriki.ru on multiple ASNs in Russia between early February 2022 to early March 2022. While OONI data shows that the site was previously accessible in Russia, we start to observe [blocking of golos-ameriki.ru](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU) from 4th March 2022 onwards. On many networks, we observe the [injection of a RST packet](https://explorer.ooni.org/measurement/20220306T142058Z_webconnectivity_RU_44943_n1_1BuqOmMQy0KMhnp7?input=https%3A%2F%2Fwww.golos-ameriki.ru%2F) after the ClientHello during the TLS handshake, while on other networks, we observe a [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220306T034232Z_webconnectivity_RU_31163_n1_8ng4wgrawcikw9qp?input=https%3A%2F%2Fwww.golos-ameriki.ru%2F). The block [varies](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-07&failure=false&domain=www.golos-ameriki.ru&probe_cc=RU) across networks, and the site remains accessible on some networks. ##### Current Time TV -[Current Time](https://www.currenttime.tv/about) is a Russian-language TV channel created by [Radio Free Europe/Radio Liberty (RFE/RL)](https://www.rferl.org/) with the participation of [Voice of America](https://www.voanews.com/). On 27th February 2022, it was [reported](https://www.interfax.ru/russia/824983) that Current Time TV received a notification from Roskomnadzor about the blocking of their website. In response, Current Time TV published [information on how to circumvent the blocking](https://d3olbu2cl74039.cloudfront.net/block) of their website. +[Current Time](https://www.currenttime.tv/about) is a Russian-language TV channel created by [Radio Free Europe/Radio Liberty (RFE/RL)](https://www.rferl.org/) with the participation of [Voice of America](https://www.voanews.com/). On 27th February 2022, it was [reported](https://www.interfax.ru/russia/824983) that Current Time TV received a notification from Roskomnadzor about the blocking of their website. In response, Current Time TV published [information on how to circumvent the blocking](https://d3olbu2cl74039.cloudfront.net/block) of their website. From 27th February 2022 onwards, OONI data suggests increased interference with access to currenttime.tv in Russia, as illustrated through the following chart. ![](images/image29.png) -Chart: OONI data on the testing of currenttime.tv in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.currenttime.tv&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.currenttime.tv&probe_cc=RU)  +**Chart:** OONI data on the testing of currenttime.tv in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.currenttime.tv&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.currenttime.tv&probe_cc=RU) -Based on our heuristics, the blocking of currenttime.tv is [automatically confirmed](https://explorer.ooni.org/search?since=2022-02-02&until=2022-03-05&failure=false&domain=www.currenttime.tv&probe_cc=RU&only=confirmed) on 2 networks in Russia ([AS44507](https://explorer.ooni.org/measurement/20220302T073622Z_webconnectivity_RU_44507_n1_buL9JKgTiQGWpbIK?input=https%3A%2F%2Fwww.currenttime.tv) and [AS212614](https://explorer.ooni.org/measurement/20220301T150906Z_webconnectivity_RU_212614_n1_7P7s1onMrMLkj2ui?input=https%3A%2F%2Fwww.currenttime.tv%2F)), as OONI data shows that the DNS resolution leads to the  IP `62.33.207.196`, which hosted the following Russian block page: +Based on our heuristics, the blocking of currenttime.tv is [automatically confirmed](https://explorer.ooni.org/search?since=2022-02-02&until=2022-03-05&failure=false&domain=www.currenttime.tv&probe_cc=RU&only=confirmed) on 2 networks in Russia ([AS44507](https://explorer.ooni.org/measurement/20220302T073622Z_webconnectivity_RU_44507_n1_buL9JKgTiQGWpbIK?input=https%3A%2F%2Fwww.currenttime.tv) and [AS212614](https://explorer.ooni.org/measurement/20220301T150906Z_webconnectivity_RU_212614_n1_7P7s1onMrMLkj2ui?input=https%3A%2F%2Fwww.currenttime.tv%2F)), as OONI data shows that the DNS resolution leads to the IP `62.33.207.196`, which hosted the following Russian block page: ![](images/image23.png) -Image: Block page served for currenttime.tv on AS44507 and AS212614 in Russia. +**Image:** Block page served for currenttime.tv on AS44507 and AS212614 in Russia. -We are also able to confirm the blocking of currenttime.tv on other networks in Russia (such as AS60139) where OONI data [shows](https://explorer.ooni.org/measurement/20220303T120931Z_webconnectivity_RU_60139_n1_RqlkvqFzTf9cySzI?input=https%3A%2F%2Fwww.currenttime.tv%2Ftv%2Fschedule%2F92%2F) that the DNS resolution leads to the  IP `185.77.150.2`, which hosts the following Russian block page: +We are also able to confirm the blocking of currenttime.tv on other networks in Russia (such as AS60139) where OONI data [shows](https://explorer.ooni.org/measurement/20220303T120931Z_webconnectivity_RU_60139_n1_RqlkvqFzTf9cySzI?input=https%3A%2F%2Fwww.currenttime.tv%2Ftv%2Fschedule%2F92%2F) that the DNS resolution leads to the IP `185.77.150.2`, which hosts the following Russian block page: ![](images/image24.png) -Image: Block page served for currenttime.tv on AS60139 network in Russia. +**Image:** Block page served for currenttime.tv on AS60139 network in Russia. -Access to currenttime.tv appears to be [blocked](https://explorer.ooni.org/search?since=2022-01-31&until=2022-03-05&failure=false&domain=www.currenttime.tv&probe_cc=RU&only=anomalies) on many other networks in Russia as well, as OONI data shows [connections being reset](https://explorer.ooni.org/measurement/20220302T111402Z_webconnectivity_RU_31364_n1_hDpBow5xK4oXoQ8G?input=https%3A%2F%2Fwww.currenttime.tv%2Ftv%2Fschedule%2F92%2F) and [timing out](https://explorer.ooni.org/measurement/20220302T112017Z_webconnectivity_RU_50716_n1_Inf9Y6D3rzAuZ5HW?input=https%3A%2F%2Fwww.currenttime.tv%2F). It’s worth highlighting though that currenttime.tv still appears to be [accessible](https://explorer.ooni.org/search?since=2022-01-31&until=2022-03-05&failure=false&domain=www.currenttime.tv&probe_cc=RU) on some networks in Russia. The OONI Measurement Aggregation Toolkit provides a [per-ASN breakdown](https://explorer.ooni.org/experimental/mat?probe_cc=RU&test_name=web_connectivity&input=https%3A%2F%2Fwww.currenttime.tv%2F&since=2022-01-31&until=2022-03-05&axis_x=measurement_start_day&axis_y=probe_asn) on the testing of currenttime.tv in Russia. +Access to currenttime.tv appears to be [blocked](https://explorer.ooni.org/search?since=2022-01-31&until=2022-03-05&failure=false&domain=www.currenttime.tv&probe_cc=RU&only=anomalies) on many other networks in Russia as well, as OONI data shows [connections being reset](https://explorer.ooni.org/measurement/20220302T111402Z_webconnectivity_RU_31364_n1_hDpBow5xK4oXoQ8G?input=https%3A%2F%2Fwww.currenttime.tv%2Ftv%2Fschedule%2F92%2F) and [timing out](https://explorer.ooni.org/measurement/20220302T112017Z_webconnectivity_RU_50716_n1_Inf9Y6D3rzAuZ5HW?input=https%3A%2F%2Fwww.currenttime.tv%2F). It’s worth highlighting though that currenttime.tv still appears to be [accessible](https://explorer.ooni.org/search?since=2022-01-31&until=2022-03-05&failure=false&domain=www.currenttime.tv&probe_cc=RU) on some networks in Russia. The OONI Measurement Aggregation Toolkit provides a [per-ASN breakdown](https://explorer.ooni.org/experimental/mat?probe_cc=RU&test_name=web_connectivity&input=https%3A%2F%2Fwww.currenttime.tv%2F&since=2022-01-31&until=2022-03-05&axis_x=measurement_start_day&axis_y=probe_asn) on the testing of currenttime.tv in Russia. ##### Meduza -[Meduza](https://meduza.io/) is a Russian-language independent news website, based out of Latvia. On 4th March 2022, Meduza [reported](https://meduza.io/feature/2022/03/04/meduza-zablokirovana-v-rossii-my-byli-k-etomu-gotovy-i-prodolzhaem-rabotat) that their website was blocked in Russia, while sharing information on how to circumvent the block. Both the [blocking of meduza.io](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-05&failure=false&domain=meduza.io&probe_cc=RU) in Russia and the timing of the block are corroborated by OONI data, as illustrated below. +[Meduza](https://meduza.io/) is a Russian-language independent news website, based out of Latvia. On 4th March 2022, Meduza [reported](https://meduza.io/feature/2022/03/04/meduza-zablokirovana-v-rossii-my-byli-k-etomu-gotovy-i-prodolzhaem-rabotat) that their website was blocked in Russia, while sharing information on how to circumvent the block. Both the [blocking of meduza.io](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-05&failure=false&domain=meduza.io&probe_cc=RU) in Russia and the timing of the block are corroborated by OONI data, as illustrated below. ![](images/image10.png) -Chart: OONI data on the testing of meduza.io in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=meduza.io&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=meduza.io&probe_cc=RU)  +**Chart:** OONI data on the testing of meduza.io in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=meduza.io&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=meduza.io&probe_cc=RU) -OONI data shows that multiple ISPs in Russia started blocking access to meduza.io on 4th March 2022, and the block remains ongoing. The block [varies](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=meduza.io&probe_cc=RU) across networks in Russia, while the site still remains accessible on a few networks. +OONI data shows that multiple ISPs in Russia started blocking access to meduza.io on 4th March 2022, and the block remains ongoing. The block [varies](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=meduza.io&probe_cc=RU) across networks in Russia, while the site still remains accessible on a few networks. ##### New Times -The [New Times](https://newtimes.ru/) is an independent Russian news magazine which was [reportedly blocked](https://www.themoscowtimes.com/2022/02/28/russia-blocks-number-of-independent-and-ukrainian-media-outlets-a76649) in Russia on 28th February 2022 for its coverage on the invasion of Ukraine. We only heard of this block on 4th March 2022, at which point we realized that newtimes.ru wasn’t already included in the [Citizen Lab test list for Russia](https://github.com/citizenlab/test-lists/blob/master/lists/ru.csv) (and, therefore, hadn’t been tested by [OONI Probe](https://ooni.org/install/) users in Russia yet). +The [New Times](https://newtimes.ru/) is an independent Russian news magazine which was [reportedly blocked](https://www.themoscowtimes.com/2022/02/28/russia-blocks-number-of-independent-and-ukrainian-media-outlets-a76649) in Russia on 28th February 2022 for its coverage on the invasion of Ukraine. We only heard of this block on 4th March 2022, at which point we realized that newtimes.ru wasn’t already included in the [Citizen Lab test list for Russia](https://github.com/citizenlab/test-lists/blob/master/lists/ru.csv) (and, therefore, hadn’t been tested by [OONI Probe](https://ooni.org/install/) users in Russia yet). -We immediately [added](https://github.com/citizenlab/test-lists/pull/930) newtimes.ru to the Russian test list and within 15 minutes, many OONI measurements on the testing of newtimes.ru had been collected and [openly published](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-05&failure=false&domain=newtimes.ru&probe_cc=RU) from several ASNs in Russia (the speed of testing was enabled by [OONI’s smart URL list system](https://ooni.org/post/ooni-smart-url-list-system/) which prioritizes the testing of recently merged URLs). The testing coverage for newtimes.ru therefore begins on 4th March 2022, when we see that a large volume of measurements (collected from multiple ASNs in Russia) suggest blocking, as illustrated below. +We immediately [added](https://github.com/citizenlab/test-lists/pull/930) newtimes.ru to the Russian test list and within 15 minutes, many OONI measurements on the testing of newtimes.ru had been collected and [openly published](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-05&failure=false&domain=newtimes.ru&probe_cc=RU) from several ASNs in Russia (the speed of testing was enabled by [OONI’s smart URL list system](https://ooni.org/post/ooni-smart-url-list-system/) which prioritizes the testing of recently merged URLs). The testing coverage for newtimes.ru therefore begins on 4th March 2022, when we see that a large volume of measurements (collected from multiple ASNs in Russia) suggest blocking, as illustrated below. ![](images/image8.png) -Chart: OONI data on the testing of newtimes.ru in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=newtimes.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=newtimes.ru&probe_cc=RU)  +**Chart:** OONI data on the testing of newtimes.ru in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=newtimes.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=newtimes.ru&probe_cc=RU) -Similarly to the testing of other blocked websites (as discussed in this report), we observe that the [blocking of newtimes.ru varies across networks](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=newtimes.ru&probe_cc=RU) in Russia (while still being accessible on a few tested networks). We primarily observe that the blocking of newtimes.ru occurs through the [injection of an RST packet](https://explorer.ooni.org/measurement/20220304T232821Z_webconnectivity_RU_8359_n1_SSowISKoztyyjFst?input=https%3A%2F%2Fnewtimes.ru%2F) following the ClientHello during the TLS handshake, but we also observe [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220304T232901Z_webconnectivity_RU_31224_n1_Be3goGoFuQ0u39R1?input=https%3A%2F%2Fnewtimes.ru%2F) and other censorship techniques. +Similarly to the testing of other blocked websites (as discussed in this report), we observe that the [blocking of newtimes.ru varies across networks](https://explorer.ooni.org/search?since=2022-03-04&until=2022-03-07&failure=false&domain=newtimes.ru&probe_cc=RU) in Russia (while still being accessible on a few tested networks). We primarily observe that the blocking of newtimes.ru occurs through the [injection of an RST packet](https://explorer.ooni.org/measurement/20220304T232821Z_webconnectivity_RU_8359_n1_SSowISKoztyyjFst?input=https%3A%2F%2Fnewtimes.ru%2F) following the ClientHello during the TLS handshake, but we also observe [TLS man-in-the-middle](https://explorer.ooni.org/measurement/20220304T232901Z_webconnectivity_RU_31224_n1_Be3goGoFuQ0u39R1?input=https%3A%2F%2Fnewtimes.ru%2F) and other censorship techniques. ##### TV Rain -TV Rain (also known as “Dozhd”) is an independent Russian television channel. On 1st March 2022, access to their website (tvrain.ru) was [reportedly blocked](https://www.newyorker.com/news/dispatch/russia-blocks-its-last-independent-television-channel) in Russia in response to their coverage of the invasion of Ukraine. Both the [blocking of tvrain.ru](https://explorer.ooni.org/search?since=2022-03-01&until=2022-03-02&failure=false&domain=tvrain.ru&probe_cc=RU) and the timing of the block are corroborated by OONI data, as illustrated below. +TV Rain (also known as “Dozhd”) is an independent Russian television channel. On 1st March 2022, access to their website (tvrain.ru) was [reportedly blocked](https://www.newyorker.com/news/dispatch/russia-blocks-its-last-independent-television-channel) in Russia in response to their coverage of the invasion of Ukraine. Both the [blocking of tvrain.ru](https://explorer.ooni.org/search?since=2022-03-01&until=2022-03-02&failure=false&domain=tvrain.ru&probe_cc=RU) and the timing of the block are corroborated by OONI data, as illustrated below. ![](images/image12.png) -Chart: OONI data on the testing of tvrain.ru in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=tvrain.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=tvrain.ru&probe_cc=RU)  +**Chart:** OONI data on the testing of tvrain.ru in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=tvrain.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=tvrain.ru&probe_cc=RU) -Starting from 1st March 2022, OONI data shows the blocking of tvrain.ru on several networks in Russia. Similarly to the testing of other blocked websites (as discussed in this report), we observe that the [blocking of tvrain.ru](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=tvrain.ru&probe_cc=RU) varies across networks in Russia (while still being accessible on a few tested networks). We primarily observe that the blocking of tvrain.ru occurs through the [injection of an RST packet](https://explorer.ooni.org/measurement/20220306T155905Z_webconnectivity_RU_44943_n1_CDgUljZIR0Wpho0S?input=https%3A%2F%2Ftvrain.ru%2F) following the ClientHello during the TLS handshake. On 3rd March 2022, TV Rain [reportedly](https://www.svoboda.org/a/v-rossii-nachali-blokirovatj-sayty-meduzy-bbc-i-deutsche-welle/31735721.html) announced the suspension of their work. +Starting from 1st March 2022, OONI data shows the blocking of tvrain.ru on several networks in Russia. Similarly to the testing of other blocked websites (as discussed in this report), we observe that the [blocking of tvrain.ru](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=tvrain.ru&probe_cc=RU) varies across networks in Russia (while still being accessible on a few tested networks). We primarily observe that the blocking of tvrain.ru occurs through the [injection of an RST packet](https://explorer.ooni.org/measurement/20220306T155905Z_webconnectivity_RU_44943_n1_CDgUljZIR0Wpho0S?input=https%3A%2F%2Ftvrain.ru%2F) following the ClientHello during the TLS handshake. On 3rd March 2022, TV Rain [reportedly](https://www.svoboda.org/a/v-rossii-nachali-blokirovatj-sayty-meduzy-bbc-i-deutsche-welle/31735721.html) announced the suspension of their work. ##### Interfax News -[Interfax](https://www.interfax.ru/) is an independent Russian news agency based in Moscow. OONI data shows that the [blocking of interfax.ru in Russia began on 26th February 2022](https://explorer.ooni.org/search?since=2022-02-26&until=2022-02-27&failure=false&domain=www.interfax.ru&probe_cc=RU). +[Interfax](https://www.interfax.ru/) is an independent Russian news agency based in Moscow. OONI data shows that the [blocking of interfax.ru in Russia began on 26th February 2022](https://explorer.ooni.org/search?since=2022-02-26&until=2022-02-27&failure=false&domain=www.interfax.ru&probe_cc=RU). -The following chart aggregates OONI measurement coverage on the testing of interfax.ru on multiple ASNs in Russia between early February 2022 to early March 2022, [showing](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.interfax.ru&probe_cc=RU) that the block began on 26th February 2020 and is currently ongoing. +The following chart aggregates OONI measurement coverage on the testing of interfax.ru on multiple ASNs in Russia between early February 2022 to early March 2022, [showing](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-07&failure=false&domain=www.interfax.ru&probe_cc=RU) that the block began on 26th February 2020 and is currently ongoing. ![](images/image18.png) -Chart: OONI data on the testing of interfax.ru in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.interfax.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.interfax.ru&probe_cc=RU)  +**Chart:** OONI data on the testing of interfax.ru in Russia, [https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.interfax.ru&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-03&until=2022-03-06&failure=false&domain=www.interfax.ru&probe_cc=RU) Similarly to the blocking of other websites (discussed in this report), the blocking of interfax.ru varies across networks in Russia, but still remains accessible on a few tested networks. #### Blocked website about captured and killed Russian soldiers -On 27th February 2022, the Ministry of Internal Affairs of Ukraine [launched](https://www.pravda.com.ua/eng/news/2022/02/27/7326424/) a new website (200rf.com) to enable Russians to find their family members who were captured or killed during the invasion of Ukraine. This website shares videos about captured and killed Russian soldiers in Ukraine, and also includes a Russian-language form that can be submitted to search for prisoners. +On 27th February 2022, the Ministry of Internal Affairs of Ukraine [launched](https://www.pravda.com.ua/eng/news/2022/02/27/7326424/) a new website (200rf.com) to enable Russians to find their family members who were captured or killed during the invasion of Ukraine. This website shares videos about captured and killed Russian soldiers in Ukraine, and also includes a Russian-language form that can be submitted to search for prisoners. -On the same day, Russia [reportedly blocked](https://www.pravda.com.ua/eng/news/2022/02/27/7326633/) access to 200rf.com. We only heard of this on 3rd March 2022, at which point we checked OONI data to see if we could confirm the blocking of 200rf.com in Russia. We realized that 200rf.com hadn’t been tested in the country yet, because it wasn’t included in the [lists of websites](https://ooni.org/support/faq/#which-websites-will-i-test-for-censorship-with-ooni-probe) tested by [OONI Probe](https://ooni.org/install/) users in Russia. We immediately [added](https://github.com/citizenlab/test-lists/pull/929) 200rf.com to the [Citizen Lab test list for Russia](https://github.com/citizenlab/test-lists/blob/master/lists/ru.csv), and OONI measurements were collected and [openly published](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU) soon thereafter (as [OONI’s smart URL list system](https://ooni.org/post/ooni-smart-url-list-system/) prioritizes the testing of recently merged URLs). +On the same day, Russia [reportedly blocked](https://www.pravda.com.ua/eng/news/2022/02/27/7326633/) access to 200rf.com. We only heard of this on 3rd March 2022, at which point we checked OONI data to see if we could confirm the blocking of 200rf.com in Russia. We realized that 200rf.com hadn’t been tested in the country yet, because it wasn’t included in the [lists of websites](https://ooni.org/support/faq/#which-websites-will-i-test-for-censorship-with-ooni-probe) tested by [OONI Probe](https://ooni.org/install/) users in Russia. We immediately [added](https://github.com/citizenlab/test-lists/pull/929) 200rf.com to the [Citizen Lab test list for Russia](https://github.com/citizenlab/test-lists/blob/master/lists/ru.csv), and OONI measurements were collected and [openly published](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU) soon thereafter (as [OONI’s smart URL list system](https://ooni.org/post/ooni-smart-url-list-system/) prioritizes the testing of recently merged URLs). -OONI data collected from multiple networks in Russia (on 3rd March 2022) corroborates the [blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-06&failure=false&domain=200rf.com&probe_cc=RU) of 200rf.com, as illustrated in the following chart.   +OONI data collected from multiple networks in Russia (on 3rd March 2022) corroborates the [blocking](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-06&failure=false&domain=200rf.com&probe_cc=RU) of 200rf.com, as illustrated in the following chart.   ![](images/image7.png) -Chart: OONI data on the testing of 200rf.com in Russia on 3rd March 2022. +**Chart:** OONI data on the testing of 200rf.com in Russia on 3rd March 2022. From the above chart, it is evident that 200rf.com was: @@ -705,29 +705,29 @@ From the above chart, it is evident that 200rf.com was: * Blocked in many different ways (i.e. we observe different censorship techniques on different networks); * Still accessible (on 3rd March 2022) on some networks in Russia. -Notably, we observe that access to 200rf.com is being interfered with through a TLS man-in-the-middle on the following 13 networks in Russia: [AS15640 (MTS PJSC)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS15640), [AS25159 (PJSC MegaFon)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS25159), [AS25513 (PJSC Moscow city telephone network)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS25513), [AS31163 (PJSC MegaFon)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS31163), [AS35807 (SkyNet Ltd.)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS35807), [AS42511 (Connect LLC)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS42511), [AS8359 (MTS PJSC)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS8359), [AS25086](https://explorer.ooni.org/measurement/20220303T215229Z_webconnectivity_RU_25086_n1_k3kIlSnoMVyUqosa?input=https%3A%2F%2F200rf.com%2F), [AS8580](https://explorer.ooni.org/measurement/20220303T214034Z_webconnectivity_RU_8580_n1_PpttWGZRuR8eD9hS?input=https%3A%2F%2F200rf.com%2F), [AS12737](https://explorer.ooni.org/measurement/20220303T212121Z_webconnectivity_RU_12737_n1_DGKkNb1wSAd9vNmo?input=https%3A%2F%2F200rf.com%2F), [AS20632](https://explorer.ooni.org/measurement/20220303T210357Z_webconnectivity_RU_20632_n1_MXDLRcv2gIdKeQG6?input=https%3A%2F%2F200rf.com%2F), [AS31286](https://explorer.ooni.org/measurement/20220303T162827Z_webconnectivity_RU_31286_n1_Ko83GT1jdv0FYH2E?input=https%3A%2F%2F200rf.com%2F), [AS31213](https://explorer.ooni.org/measurement/20220303T120016Z_webconnectivity_RU_31213_n1_MI62mKfi2ZFeiC6R?input=https%3A%2F%2F200rf.com%2F). +Notably, we observe that access to 200rf.com is being interfered with through a TLS man-in-the-middle on the following 13 networks in Russia: [AS15640 (MTS PJSC)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS15640), [AS25159 (PJSC MegaFon)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS25159), [AS25513 (PJSC Moscow city telephone network)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS25513), [AS31163 (PJSC MegaFon)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS31163), [AS35807 (SkyNet Ltd.)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS35807), [AS42511 (Connect LLC)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS42511), [AS8359 (MTS PJSC)](https://explorer.ooni.org/search?since=2022-02-01&until=2022-03-04&failure=false&domain=200rf.com&probe_cc=RU&probe_asn=AS8359), [AS25086](https://explorer.ooni.org/measurement/20220303T215229Z_webconnectivity_RU_25086_n1_k3kIlSnoMVyUqosa?input=https%3A%2F%2F200rf.com%2F), [AS8580](https://explorer.ooni.org/measurement/20220303T214034Z_webconnectivity_RU_8580_n1_PpttWGZRuR8eD9hS?input=https%3A%2F%2F200rf.com%2F), [AS12737](https://explorer.ooni.org/measurement/20220303T212121Z_webconnectivity_RU_12737_n1_DGKkNb1wSAd9vNmo?input=https%3A%2F%2F200rf.com%2F), [AS20632](https://explorer.ooni.org/measurement/20220303T210357Z_webconnectivity_RU_20632_n1_MXDLRcv2gIdKeQG6?input=https%3A%2F%2F200rf.com%2F), [AS31286](https://explorer.ooni.org/measurement/20220303T162827Z_webconnectivity_RU_31286_n1_Ko83GT1jdv0FYH2E?input=https%3A%2F%2F200rf.com%2F), [AS31213](https://explorer.ooni.org/measurement/20220303T120016Z_webconnectivity_RU_31213_n1_MI62mKfi2ZFeiC6R?input=https%3A%2F%2F200rf.com%2F). -On other networks (such as AS8790), OONI data [shows](https://explorer.ooni.org/measurement/20220303T111548Z_webconnectivity_RU_8790_n1_rMZoIX9SJJI0NRB0?input=https%3A%2F%2F200rf.com%2F) that access to 200rf.com is blocked by means of DNS tampering, as the DNS queries resolve to the IP `85.142.29.248` which hosts the following Russian blockpage: +On other networks (such as AS8790), OONI data [shows](https://explorer.ooni.org/measurement/20220303T111548Z_webconnectivity_RU_8790_n1_rMZoIX9SJJI0NRB0?input=https%3A%2F%2F200rf.com%2F) that access to 200rf.com is blocked by means of DNS tampering, as the DNS queries resolve to the IP `85.142.29.248` which hosts the following Russian blockpage: ![](images/image6.png) -Image: Block page served for 200rf.com on AS8790 in Russia. +**Image:** Block page served for 200rf.com on AS8790 in Russia. -On other networks (such as [AS60139](https://explorer.ooni.org/measurement/20220303T111938Z_webconnectivity_RU_60139_n1_o7KFBFqAwSFdVcW4?input=https%3A%2F%2F200rf.com%2F) and [AS12389](https://explorer.ooni.org/measurement/20220303T111841Z_webconnectivity_RU_12389_n1_TNulGfMNA5nfH7P4?input=https%3A%2F%2F200rf.com%2F)), OONI data shows the injection of a RST packet during the TLS handshake. While on other networks, we [see](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-06&failure=false&domain=200rf.com&probe_cc=RU&only=anomalies) that the session times out during the TLS handshake, or that the connection is closed before the TLS handshake is complete. +On other networks (such as [AS60139](https://explorer.ooni.org/measurement/20220303T111938Z_webconnectivity_RU_60139_n1_o7KFBFqAwSFdVcW4?input=https%3A%2F%2F200rf.com%2F) and [AS12389](https://explorer.ooni.org/measurement/20220303T111841Z_webconnectivity_RU_12389_n1_TNulGfMNA5nfH7P4?input=https%3A%2F%2F200rf.com%2F)), OONI data shows the injection of a RST packet during the TLS handshake. While on other networks, we [see](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-06&failure=false&domain=200rf.com&probe_cc=RU&only=anomalies) that the session times out during the TLS handshake, or that the connection is closed before the TLS handshake is complete. #### Twitter and Facebook blocked -On 26th February 2022, we started to observe anomalies in the testing of twitter.com on some networks in Russia. Those [measurements suggested potential throttling of Twitter](https://explorer.ooni.org/search?probe_cc=RU&test_name=web_connectivity&since=2022-01-26&until=2022-02-27&failure=false&domain=twitter.com) because the `network_event` keys of raw OONI data showed that the TLS handshake was performed by reading very small chunks of data very slowly, sometimes leading to a timeout. This pattern changed by 4th March 2022, when we instead started to primarily observe blocking of twitter.com through the [injection of a RST packet](https://explorer.ooni.org/measurement/20220304T234236Z_webconnectivity_RU_12389_n1_Sy7B8SjkD18aFR3r?input=https%3A%2F%2Ftwitter.com%2F) following the ClientHello during the TLS handshake. +On 26th February 2022, we started to observe anomalies in the testing of twitter.com on some networks in Russia. Those [measurements suggested potential throttling of Twitter](https://explorer.ooni.org/search?probe_cc=RU&test_name=web_connectivity&since=2022-01-26&until=2022-02-27&failure=false&domain=twitter.com) because the `network_event` keys of raw OONI data showed that the TLS handshake was performed by reading very small chunks of data very slowly, sometimes leading to a timeout. This pattern changed by 4th March 2022, when we instead started to primarily observe blocking of twitter.com through the [injection of a RST packet](https://explorer.ooni.org/measurement/20220304T234236Z_webconnectivity_RU_12389_n1_Sy7B8SjkD18aFR3r?input=https%3A%2F%2Ftwitter.com%2F) following the ClientHello during the TLS handshake. -Starting from 3rd March 2022, OONI data [shows](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-04&failure=false&domain=www.facebook.com&probe_cc=RU) the blocking of facebook.com on multiple networks in Russia, where we also observe [connection resets](https://explorer.ooni.org/measurement/20220304T234236Z_webconnectivity_RU_12389_n1_Sy7B8SjkD18aFR3r?input=https%3A%2F%2Fwww.facebook.com%2F) (similarly to the blocking of Twitter). Roskomnadzor published a [statement](https://rkn.gov.ru/news/rsoc/news74156.htm) disclosing their decision to block Facebook. They argue that they have recorded 26 cases of discrimination against Russian media and information resources by Facebook, mentioning that Facebook recently restricted access to several Russian news agency accounts (such as RT and Sputnik). Both facebook.com and twitter.com have since been added to [Roskomnadzor’s blocklist](https://blocklist.rkn.gov.ru/), where it is disclosed that they are blocked on the grounds of violating [Article 15.3](https://blocklist.rkn.gov.ru/article15-3/) of the Federal Law.  +Starting from 3rd March 2022, OONI data [shows](https://explorer.ooni.org/search?since=2022-03-03&until=2022-03-04&failure=false&domain=www.facebook.com&probe_cc=RU) the blocking of facebook.com on multiple networks in Russia, where we also observe [connection resets](https://explorer.ooni.org/measurement/20220304T234236Z_webconnectivity_RU_12389_n1_Sy7B8SjkD18aFR3r?input=https%3A%2F%2Fwww.facebook.com%2F) (similarly to the blocking of Twitter). Roskomnadzor published a [statement](https://rkn.gov.ru/news/rsoc/news74156.htm) disclosing their decision to block Facebook. They argue that they have recorded 26 cases of discrimination against Russian media and information resources by Facebook, mentioning that Facebook recently restricted access to several Russian news agency accounts (such as RT and Sputnik). Both facebook.com and twitter.com have since been added to [Roskomnadzor’s blocklist](https://blocklist.rkn.gov.ru/), where it is disclosed that they are blocked on the grounds of violating [Article 15.3](https://blocklist.rkn.gov.ru/article15-3/) of the Federal Law. The following chart aggregates OONI measurement coverage from Russia, showing that the blocking of facebook.com started several days after the blocking of twitter.com. ![](images/image5.png) -Chart: OONI data on the testing of Twitter and Facebook domains in Russia (February - March 2022), [https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU), [https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU)  +**Chart:** OONI data on the testing of Twitter and Facebook domains in Russia (February - March 2022), [https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=twitter.com&probe_cc=RU), [https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU) -Similarly to the blocking of news media websites (discussed in previous sections of this report), OONI data shows that access to [facebook.com is blocked](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies) in Russia through the use of similar censorship techniques. +Similarly to the blocking of news media websites (discussed in previous sections of this report), OONI data shows that access to [facebook.com is blocked](https://explorer.ooni.org/search?since=2022-02-04&until=2022-03-07&failure=false&domain=www.facebook.com&probe_cc=RU&only=anomalies) in Russia through the use of similar censorship techniques. The following table summarizes the methods we have observed on 56 different networks in Russia for which we have a strong indication of facebook.com blocking. @@ -861,7 +861,7 @@ Below we provide a scatterplot of this metric over time grouped by ASN. ![](images/image2.png) -Through the above chart we can see that there was a change in the blocking pattern for twitter.com between 26th February 2022 to 4th March 2022. In particular, we can see that between these dates, the `read_speed` metric (for many samples) during the TLS handshake for twitter.com is very slow (suggesting Twitter throttling), while from 4th March 2022 onwards, we observe that the `read_speed` metric resumes to previous levels. However, [OONI data](https://explorer.ooni.org/search?probe_cc=RU&test_name=web_connectivity&since=2022-03-03&until=2022-03-07&failure=false&domain=twitter.com&only=anomalies) from 4th March 2022 onwards shows [connection resets](https://explorer.ooni.org/measurement/20220306T173949Z_webconnectivity_RU_12389_n1_91bgJ29ePo9jdHyk?input=https%3A%2F%2Ftwitter.com%2FGraniTweet%2F), suggesting that interference to twitter.com **changed from throttling to blocking** (through the injection of a RST packet). +Through the above chart we can see that there was a change in the blocking pattern for twitter.com between 26th February 2022 to 4th March 2022. In particular, we can see that between these dates, the `read_speed` metric (for many samples) during the TLS handshake for twitter.com is very slow (suggesting Twitter throttling), while from 4th March 2022 onwards, we observe that the `read_speed` metric resumes to previous levels. However, [OONI data](https://explorer.ooni.org/search?probe_cc=RU&test_name=web_connectivity&since=2022-03-03&until=2022-03-07&failure=false&domain=twitter.com&only=anomalies) from 4th March 2022 onwards shows [connection resets](https://explorer.ooni.org/measurement/20220306T173949Z_webconnectivity_RU_12389_n1_91bgJ29ePo9jdHyk?input=https%3A%2F%2Ftwitter.com%2FGraniTweet%2F), suggesting that interference to twitter.com **changed from throttling to blocking** (through the injection of a RST packet). As another proxy for inferring unusual behavior at the network level during the TLS handshake, we also plotted the number of distinct read operations over time, as illustrated below. We consider this to be a good enough proxy, because we make the assumption that the amount of data downloaded during a TLS handshake is constant. @@ -869,17 +869,17 @@ As another proxy for inferring unusual behavior at the network level during the The idea behind this is that if we have recorded many distinct reads, it’s a sign that data is being read in small chunks and is likely the result of some form of throttling. Again, we notice the same pattern of metrics resuming to normal levels on 4th March 2022. -What is also interesting to note is that the **throttling of twitter.com seems to stop across all ISPs in Russia at the same time** (~08:00 on 4th March 2022 UTC), as illustrated below. +What is also interesting to note is that the **throttling of twitter.com seems to stop across all ISPs in Russia at the same time** (~08:00 on 4th March 2022 UTC), as illustrated below. ![](images/image4.png) -This may suggest that throttling is performed in some sort of centralized way, as opposed to relying on ISPs to implement it independently. This finding is consistent with what was recently documented in a [research paper](https://censoredplanet.org/assets/throttling-imc-paper.pdf) on Twitter throttling in Russia.   +This may suggest that throttling is performed in some sort of centralized way, as opposed to relying on ISPs to implement it independently. This finding is consistent with what was recently documented in a [research paper](https://censoredplanet.org/assets/throttling-imc-paper.pdf) on Twitter throttling in Russia.   Furthermore, we break down the `read_speed` metric by destination endpoint to conclude that there doesn’t seem to be any difference in the endpoint being contacted when performing the TLS handshake. ![](images/image31.png) -It’s worth highlighting that we do not see a similar pattern in the testing of facebook.com (which, as discussed in the previous section, is blocked through other means). +It’s worth highlighting that we do not see a similar pattern in the testing of facebook.com (which, as discussed in the previous section, is blocked through other means). ![](images/image14.png) @@ -887,19 +887,19 @@ The above graph (based on the `read_speed` metric during the TLS handshake for f ### EU -Following the recent [decision of the Council of the European Union](https://www.consilium.europa.eu/en/press/press-releases/2022/03/02/eu-imposes-sanctions-on-state-owned-outlets-rt-russia-today-and-sputnik-s-broadcasting-in-the-eu/) to suspend the broadcasting activities of Sputnik and Russia Today (RT) in the EU, some EU countries have started to implement blocks. In Greece, for example, OONI measurements collected from the [Vodafone network (AS3329)](https://explorer.ooni.org/measurement/20220304T144806Z_webconnectivity_GR_3329_n1_dsFVoQ4VJpoCt8fI?input=https%3A%2F%2Fwww.rt.com) show an IP hosting the following block page in response to DNS queries for rt.com: +Following the recent [decision of the Council of the European Union](https://www.consilium.europa.eu/en/press/press-releases/2022/03/02/eu-imposes-sanctions-on-state-owned-outlets-rt-russia-today-and-sputnik-s-broadcasting-in-the-eu/) to suspend the broadcasting activities of Sputnik and Russia Today (RT) in the EU, some EU countries have started to implement blocks. In Greece, for example, OONI measurements collected from the [Vodafone network (AS3329)](https://explorer.ooni.org/measurement/20220304T144806Z_webconnectivity_GR_3329_n1_dsFVoQ4VJpoCt8fI?input=https%3A%2F%2Fwww.rt.com) show an IP hosting the following block page in response to DNS queries for rt.com: ![](images/image25.png) -Image: Block page served for rt.com on Vodafone (AS3329) in Greece. +**Image:** Block page served for rt.com on Vodafone (AS3329) in Greece. -The reason for the block is stated to be related to copyright infringement, though that seems unlikely in this case. Similarly, OONI data collected from France shows that rt.com is blocked on [SFR (AS15557)](https://explorer.ooni.org/measurement/20220306T025735Z_webconnectivity_FR_15557_n1_KvMNzvez6OP3omZT?input=https%3A%2F%2Fwww.rt.com%2F), as DNS queries resolve to an IP that hosts a block page (displaying a message related to customer service). We also observe blocking of rt.com in several other EU countries, such as [Romania](https://explorer.ooni.org/measurement/20220304T030739Z_webconnectivity_RO_8708_n1_tKowJ4mIheQGZPiz?input=https%3A%2F%2Fwww.rt.com%2F), [Spain](https://explorer.ooni.org/measurement/20220304T101513Z_webconnectivity_ES_57269_n1_D7A7ZT1FeLjafEOE?input=https%3A%2F%2Fwww.rt.com%2F), [Poland](https://explorer.ooni.org/measurement/20220304T050514Z_webconnectivity_PL_12741_n1_bp9mc7xQI7gwVUAO?input=https%3A%2F%2Fwww.rt.com%2F), and [Germany](https://explorer.ooni.org/measurement/20220304T165036Z_webconnectivity_DE_3320_n1_DlRC00SS0DwYLpYB?input=https%3A%2F%2Fwww.rt.com%2F). Moreover, we observe the blocking of sputniknews.com in several European countries, such as [France](https://explorer.ooni.org/measurement/20220304T011137Z_webconnectivity_FR_15557_n1_mUlmsixqs8azGuO0?input=https%3A%2F%2Fsputniknews.com%2F), [Germany](https://explorer.ooni.org/measurement/20220307T100807Z_webconnectivity_DE_3320_n1_BuMiChfm9KH2yGCU?input=https%3A%2F%2Fsputniknews.com%2F), [Spain](https://explorer.ooni.org/measurement/20220307T094622Z_webconnectivity_ES_57269_n1_reCUJVZ0B5LSbJ7G?input=https%3A%2F%2Fsputniknews.com%2F), and the [Czech Republic](https://explorer.ooni.org/measurement/20220304T141642Z_webconnectivity_CZ_28725_n1_gz9A36dCD9LIPA0K?input=https%3A%2F%2Fsputniknews.com%2F).     +The reason for the block is stated to be related to copyright infringement, though that seems unlikely in this case. Similarly, OONI data collected from France shows that rt.com is blocked on [SFR (AS15557)](https://explorer.ooni.org/measurement/20220306T025735Z_webconnectivity_FR_15557_n1_KvMNzvez6OP3omZT?input=https%3A%2F%2Fwww.rt.com%2F), as DNS queries resolve to an IP that hosts a block page (displaying a message related to customer service). We also observe blocking of rt.com in several other EU countries, such as [Romania](https://explorer.ooni.org/measurement/20220304T030739Z_webconnectivity_RO_8708_n1_tKowJ4mIheQGZPiz?input=https%3A%2F%2Fwww.rt.com%2F), [Spain](https://explorer.ooni.org/measurement/20220304T101513Z_webconnectivity_ES_57269_n1_D7A7ZT1FeLjafEOE?input=https%3A%2F%2Fwww.rt.com%2F), [Poland](https://explorer.ooni.org/measurement/20220304T050514Z_webconnectivity_PL_12741_n1_bp9mc7xQI7gwVUAO?input=https%3A%2F%2Fwww.rt.com%2F), and [Germany](https://explorer.ooni.org/measurement/20220304T165036Z_webconnectivity_DE_3320_n1_DlRC00SS0DwYLpYB?input=https%3A%2F%2Fwww.rt.com%2F). Moreover, we observe the blocking of sputniknews.com in several European countries, such as [France](https://explorer.ooni.org/measurement/20220304T011137Z_webconnectivity_FR_15557_n1_mUlmsixqs8azGuO0?input=https%3A%2F%2Fsputniknews.com%2F), [Germany](https://explorer.ooni.org/measurement/20220307T100807Z_webconnectivity_DE_3320_n1_BuMiChfm9KH2yGCU?input=https%3A%2F%2Fsputniknews.com%2F), [Spain](https://explorer.ooni.org/measurement/20220307T094622Z_webconnectivity_ES_57269_n1_reCUJVZ0B5LSbJ7G?input=https%3A%2F%2Fsputniknews.com%2F), and the [Czech Republic](https://explorer.ooni.org/measurement/20220304T141642Z_webconnectivity_CZ_28725_n1_gz9A36dCD9LIPA0K?input=https%3A%2F%2Fsputniknews.com%2F). -On 4th March 2022, the [European Commission requested](https://lumendatabase.org/notices/26927483) Google to remove results from Google Search that pertain to rt.com and sputniknews.com. The following screenshot is taken from our web browser in Italy (when searching for rt.com). +On 4th March 2022, the [European Commission requested](https://lumendatabase.org/notices/26927483) Google to remove results from Google Search that pertain to rt.com and sputniknews.com. The following screenshot is taken from our web browser in Italy (when searching for rt.com). ![](images/image19.png) -Moreover, [Facebook has recently made the RTNews Facebook page unavailable](https://www.politico.eu/article/facebook-parent-company-restricts-eu-access-to-russia-rt-sputnik/) to users accessing it from Europe. We analyzed OONI measurements from 62 countries and 221 networks using a custom heuristic that is able to distinguish between a Facebook geoblock and an available page. When a specific Facebook page is geoblocked, a login page is presented when a user attempts to access it, while a different page is shown when it isn’t. +Moreover, [Facebook has recently made the RTNews Facebook page unavailable](https://www.politico.eu/article/facebook-parent-company-restricts-eu-access-to-russia-rt-sputnik/) to users accessing it from Europe. We analyzed OONI measurements from 62 countries and 221 networks using a custom heuristic that is able to distinguish between a Facebook geoblock and an available page. When a specific Facebook page is geoblocked, a login page is presented when a user attempts to access it, while a different page is shown when it isn’t. When Facebook users in Europe attempt to access the RTNews page while logged in, the following page is presented: @@ -907,7 +907,7 @@ When Facebook users in Europe attempt to access the RTNews page while logged in, Based on OONI measurements, we were able to confirm that Facebook is blocking access to the RTNews page for users in the following 19 countries: [Austria (AT)](https://explorer.ooni.org/measurement/20220306T210225Z_webconnectivity_AT_8447_n1_fwLwEdNLIj0QiSo9?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Belgium (BE)](https://explorer.ooni.org/measurement/20220306T210926Z_webconnectivity_BE_12392_n1_sG5mC5UnRykwr9Xf?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Czechia (CZ)](https://explorer.ooni.org/measurement/20220306T211233Z_webconnectivity_CZ_16019_n1_h4V2X6qXCJJyYzC1?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Germany (DE)](https://explorer.ooni.org/measurement/20220306T212241Z_webconnectivity_DE_8422_n1_cQBffpeCBU0fo6uk?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Denmark (DK)](https://explorer.ooni.org/measurement/20220306T211917Z_webconnectivity_DK_209424_n1_ToegUOFED9o9SnYM?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Spain (ES)](https://explorer.ooni.org/measurement/20220306T212412Z_webconnectivity_ES_12430_n1_gR565oHabLNjUCFo?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Finland (FI)](https://explorer.ooni.org/measurement/20220306T212812Z_webconnectivity_FI_24940_n1_SrZlX6hdzkhcBMot?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [France (FR)](https://explorer.ooni.org/measurement/20220306T212658Z_webconnectivity_FR_51207_n1_o7uEURSMoXto7Xe7?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [United Kingdom (GB)](https://explorer.ooni.org/measurement/20220306T212248Z_webconnectivity_GB_5607_n1_iezd2nkTHyyrZk4E?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Hungary (HU)](https://explorer.ooni.org/measurement/20220306T204455Z_webconnectivity_HU_213155_n1_VD1a6oo886jTbGpq?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Ireland (IE)](https://explorer.ooni.org/measurement/20220306T212211Z_webconnectivity_IE_15751_n1_D5Z5RYcOPuDYVyqu?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Italy (IT)](https://explorer.ooni.org/measurement/20220306T212618Z_webconnectivity_IT_30722_n1_XJ4OoZaLsE5zJhzW?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Netherlands (NL)](https://explorer.ooni.org/measurement/20220306T212457Z_webconnectivity_NL_13127_n1_ROmuzUyW2kdKNLsr?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Poland (PL)](https://explorer.ooni.org/measurement/20220306T210513Z_webconnectivity_PL_6830_n1_tiDLY5X4BvEbgf4n?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Portugal (PT)](https://explorer.ooni.org/measurement/20220306T205906Z_webconnectivity_PT_12353_n1_e70hrvcpiJY8Rwfw?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Sweden (SE)](https://explorer.ooni.org/measurement/20220306T212703Z_webconnectivity_SE_2119_n1_enYyM1xOyAy3I3lr?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Slovenia (SI)](https://explorer.ooni.org/measurement/20220306T205407Z_webconnectivity_SI_21283_n1_8F6bZv8vEkIv1VSM?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Slovakia (SK)](https://explorer.ooni.org/measurement/20220306T205352Z_webconnectivity_SK_197398_n1_H3gEapy67NMWkLGn?input=https%3A%2F%2Fwww.facebook.com%2FRTNews), [Ukraine (UA)](https://explorer.ooni.org/measurement/20220306T212417Z_webconnectivity_UA_34743_n1_15WMAWgpEYxVIO0V?input=https%3A%2F%2Fwww.facebook.com%2FRTNews). These OONI measurements are annotated as accessible on OONI Explorer because we don’t currently have this custom heuristic integrated into our data processing pipeline. It’s interesting to see that Facebook also blocks the RTNews page for users in countries that are not technically part of the EU, such as the UK. -Similar geoblocks have been implemented by Twitter. When attempting to access [RT_com](https://twitter.com/RT_com) from an EU country, the following page is displayed: +Similar geoblocks have been implemented by Twitter. When attempting to access [RT_com](https://twitter.com/RT_com) from an EU country, the following page is displayed: ![](images/image13.png) @@ -915,6 +915,6 @@ Similar geoblocks have been implemented by Twitter. When attempting to access [R Russia has ramped up its censorship in an attempt to control the narrative surrounding the war in Ukraine. As the implementation of internet censorship is decentralized in Russia, different internet users in Russia may experience different blocks, depending on which network they’re connected to. Yet, the seemingly centralized throttling of Twitter raises alarms about Russia potentially having centralized censorship capabilities, which would make censorship more effective and harder to circumvent. -EU countries are not known for blocking news media websites, and the recent [decision of the Council of the European Union](https://www.consilium.europa.eu/en/press/press-releases/2022/03/02/eu-imposes-sanctions-on-state-owned-outlets-rt-russia-today-and-sputnik-s-broadcasting-in-the-eu/) to suspend the broadcasting activities of Sputnik and Russia Today (RT) in the EU raises the question of whether this would be setting a precedent for further censorship in the future. +EU countries are not known for blocking news media websites, and the recent [decision of the Council of the European Union](https://www.consilium.europa.eu/en/press/press-releases/2022/03/02/eu-imposes-sanctions-on-state-owned-outlets-rt-russia-today-and-sputnik-s-broadcasting-in-the-eu/) to suspend the broadcasting activities of Sputnik and Russia Today (RT) in the EU raises the question of whether this would be setting a precedent for further censorship in the future. Both Russia and the EU argue that these new media blocks are on the grounds of limiting the spread of misinformation. But this may risk evolving into a censorship war. diff --git a/content/post/2022-state-of-internet-censorship-cambodia/index.md b/content/post/2022-state-of-internet-censorship-cambodia/index.md index 1996af5b5..04a22cb70 100644 --- a/content/post/2022-state-of-internet-censorship-cambodia/index.md +++ b/content/post/2022-state-of-internet-censorship-cambodia/index.md @@ -8,7 +8,7 @@ author: " Lam Socheat (API), Tan Thary (API), Chan Vibol (Consultant), Siti Nurl date: "2022-12-23" -tags: ["cambodia", "censorship", "country-kh"] +tags: ["cambodia", "censorship", "country-kh", "theme-news_media"] categories: ["report"] diff --git a/content/post/2022-state-of-internet-censorship-hong-kong/index.md b/content/post/2022-state-of-internet-censorship-hong-kong/index.md index 64632d22e..235da8ab5 100644 --- a/content/post/2022-state-of-internet-censorship-hong-kong/index.md +++ b/content/post/2022-state-of-internet-censorship-hong-kong/index.md @@ -8,7 +8,7 @@ author: " Independent Researchers (Anonymous), Siti Nurliza Samsudin (Sinar Proj date: "2022-12-23" -tags: ["hong-kong", "censorship", "country-hk"] +tags: ["hong-kong", "censorship", "country-hk", "theme-human_rights"] categories: ["report"] diff --git a/content/post/2022-state-of-internet-censorship-indonesia/index.md b/content/post/2022-state-of-internet-censorship-indonesia/index.md index 34999648d..126e5d185 100644 --- a/content/post/2022-state-of-internet-censorship-indonesia/index.md +++ b/content/post/2022-state-of-internet-censorship-indonesia/index.md @@ -8,7 +8,7 @@ author: "Khairil Zhafri (EngageMedia), Pradipa P. Rasidi (EngageMedia), Debby Kr date: "2022-12-23" -tags: ["indonesia", "censorship", "country-id"] +tags: ["indonesia", "censorship", "country-id", "theme-circumvention", "theme-human_rights", "theme-news_media"] categories: ["report"] diff --git a/content/post/2022-state-of-internet-censorship-malaysia/index.md b/content/post/2022-state-of-internet-censorship-malaysia/index.md index 5bea219c4..6ed03a9f7 100644 --- a/content/post/2022-state-of-internet-censorship-malaysia/index.md +++ b/content/post/2022-state-of-internet-censorship-malaysia/index.md @@ -8,7 +8,7 @@ author: "Kelly Koh (Sinar Project) and Siti Nurliza Samsudin (Sinar Project)" date: "2022-12-23" -tags: ["malaysia", "censorship", "country-my"] +tags: ["malaysia", "censorship", "country-my", "theme-news_media", "theme-human_rights"] categories: ["report"] diff --git a/content/post/2022-state-of-internet-censorship-myanmar/index.md b/content/post/2022-state-of-internet-censorship-myanmar/index.md index 15976a456..ef080fdce 100644 --- a/content/post/2022-state-of-internet-censorship-myanmar/index.md +++ b/content/post/2022-state-of-internet-censorship-myanmar/index.md @@ -8,7 +8,7 @@ author: " Independent Researcher (Anonymous), Kelly Koh (Sinar Project), Siti Nu date: "2022-12-23" -tags: ["myanmar", "censorship", "country-mm"] +tags: ["myanmar", "censorship", "country-mm", "theme-news_media", "theme-circumvention", "theme-im"] categories: ["report"] diff --git a/content/post/2022-state-of-internet-censorship-philippines/index.md b/content/post/2022-state-of-internet-censorship-philippines/index.md index d6dff8e17..eb4b7f77b 100644 --- a/content/post/2022-state-of-internet-censorship-philippines/index.md +++ b/content/post/2022-state-of-internet-censorship-philippines/index.md @@ -8,7 +8,7 @@ author: "Katerina Francisco (EngageMedia), Khairil Zhafri (EngageMedia), Red Tan date: "2022-12-23" -tags: ["philippines", "censorship", "country-ph"] +tags: ["philippines", "censorship", "country-ph", "theme-circumvention", "theme-social_media"] categories: ["report"] diff --git a/content/post/2022-state-of-internet-censorship-thailand/index.md b/content/post/2022-state-of-internet-censorship-thailand/index.md index 7cc71ccb3..6bbc9e1cf 100644 --- a/content/post/2022-state-of-internet-censorship-thailand/index.md +++ b/content/post/2022-state-of-internet-censorship-thailand/index.md @@ -8,7 +8,7 @@ author: "Siti Nurliza Samsudin (Sinar Project), Kelly Koh (Sinar Project) and Th date: "2022-12-23" -tags: ["thailand", "censorship", "country-th"] +tags: ["thailand", "censorship", "country-th", "theme-news_media", "theme-human_rights", "theme-circumvention"] categories: ["report"] @@ -18,9 +18,6 @@ Thailand, a country situated on the mainland of Southeast Asia, was under milita This report is aimed to portray the state of internet censorship in Thailand based on these reported incidents, as well as OONI measurements collected during the first half of 2022. -In this report, we share our technical findings. In both Armenia and Azerbaijan, we found TLS and DNS level interference of TikTok domains and endpoints during the border clashes. - - ## Table of Contents **[Introduction](#introduction)** diff --git a/content/post/2022-state-of-internet-censorship-vietnam/index.md b/content/post/2022-state-of-internet-censorship-vietnam/index.md index a26882694..6ca6d4fbe 100644 --- a/content/post/2022-state-of-internet-censorship-vietnam/index.md +++ b/content/post/2022-state-of-internet-censorship-vietnam/index.md @@ -8,7 +8,7 @@ author: "Independent Researchers (Anonymous), Kelly Koh (Sinar Project) and Sit date: "2022-12-23" -tags: ["vietnam", "censorship", "country-th"] +tags: ["vietnam", "censorship", "country-th", "theme-news_media", "theme-human_rights"] categories: ["report"] diff --git a/content/post/2023-azerbaijan-internet-censorship/index.md b/content/post/2023-azerbaijan-internet-censorship/index.md index 0e14a3668..1c6ae9e0d 100644 --- a/content/post/2023-azerbaijan-internet-censorship/index.md +++ b/content/post/2023-azerbaijan-internet-censorship/index.md @@ -3,76 +3,76 @@ title: "OONI measurements show ongoing internet censorship in Azerbaijan" description: "This report shares OONI censorship findings from Azerbaijan." author: "Maria Xynou (OONI), Arzu Geybulla (Azerbaijan Internet Watch)" date: "2023-02-28" -tags: ["country-az", "censorship", "azerbaijan"] +tags: ["country-az", "censorship", "azerbaijan", "theme-news_media", "theme-circumvention", "theme-social_media"] categories: ["report"] --- -Azerbaijan is known to block access to independent news media websites – we previously [reported](https://ooni.org/post/2021-azerbaijan/) on this in July 2021. At the time, we analyzed OONI measurements collected from Azerbaijan between January 2020 to May 2021 and found that ISPs in Azerbaijan were [blocking access to several independent news media](https://ooni.org/post/2021-azerbaijan/#blocked-news-media-websites) and [circumvention tool sites](https://ooni.org/post/2021-azerbaijan/#blocked-circumvention-tool-sites). We also found that amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), ISPs in Azerbaijan [temporarily blocked access to social media](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war) services, and [attempted to block](https://ooni.org/post/2021-azerbaijan/#tor-and-psiphon) access to [Tor](https://www.torproject.org/) and [Psiphon](https://psiphon.ca/). +Azerbaijan is known to block access to independent news media websites – we previously [reported](https://ooni.org/post/2021-azerbaijan/) on this in July 2021. At the time, we analyzed OONI measurements collected from Azerbaijan between January 2020 to May 2021 and found that ISPs in Azerbaijan were [blocking access to several independent news media](https://ooni.org/post/2021-azerbaijan/#blocked-news-media-websites) and [circumvention tool sites](https://ooni.org/post/2021-azerbaijan/#blocked-circumvention-tool-sites). We also found that amid the [2020 Nagorno-Karabakh war](https://www.bbc.com/news/world-europe-54324772), ISPs in Azerbaijan [temporarily blocked access to social media](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war) services, and [attempted to block](https://ooni.org/post/2021-azerbaijan/#tor-and-psiphon) access to [Tor](https://www.torproject.org/) and [Psiphon](https://psiphon.ca/). But how has Azerbaijan’s internet censorship landscape changed (if at all) over the last year? -We attempt to address this question as part of this report. Specifically, we analyzed [OONI data collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between **January 2022 to February 2023** with the goal of evaluating which tested services presented signs of blocking, and whether new blocks had emerged over the past year (in comparison to our [previous report](https://ooni.org/post/2021-azerbaijan/)). +We attempt to address this question as part of this report. Specifically, we analyzed [OONI data collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between **January 2022 to February 2023** with the goal of evaluating which tested services presented signs of blocking, and whether new blocks had emerged over the past year (in comparison to our [previous report](https://ooni.org/post/2021-azerbaijan/)). {{}} ## Key Findings -As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between January 2022 to February 2023, we found: +As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between January 2022 to February 2023, we found: -* **Blocking of news media websites.** Azerbaijan continues to [block access to several independent news media websites](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS). OONI data also suggests that some ISPs in Azerbaijan [may have started blocking access to](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) [t](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies)[he Guardian](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) on 25th December 2022. -* **Azerbaijan and Russia block each other’s news media.** In [early June 2022](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm), Azerbaijan started [blocking access to Russia’s state-run RIA Novosti](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&domain=ria.ru) media website. Since (at least) 7th June 2022, Russian ISPs have been [blocking](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-06-01&until=2023-02-25&time_grain=day&probe_cc=RU&axis_y=domain) access to Azerbaijani news media websites (`haqqin.az`, `minval.az`, `oxu.az`, `ru.oxu.az`, `ru.baku.ws`). These blocks remain ongoing. -* **Temporary blocking of TikTok amid border clashes with Armenia.** During the September 2022 border clashes, both [Azerbaijan and Armenia blocked access to TikTok](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/). While the TikTok block was [lifted](https://explorer.ooni.org/search?since=2022-09-01&until=2022-09-30&failure=false&domain=www.tiktok.com&probe_cc=AM) in Armenia by 21st September 2022 (only lasting about a week), the [TikTok block](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) remained in place in Azerbaijan until November 2022 (lasting about 2 months). -* **Blocking of circumvention tool sites.** Azerbaijan continues to [block access to numerous circumvention tool websites](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=ANON), potentially limiting the ability to circumvent internet censorship in Azerbaijan. However, most OONI measurements suggest that tested circumvention tools ([Tor](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=tor) and [Psiphon](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=psiphon)) appear to be reachable. -* **Variance of censorship across networks.** While most ISPs in Azerbaijan appear to adopt similar censorship techniques (as we continued to observe connection timeouts in most anomalous measurements across ASNs), different ISPs block access to different websites over time. +* **Blocking of news media websites.** Azerbaijan continues to [block access to several independent news media websites](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS). OONI data also suggests that some ISPs in Azerbaijan [may have started blocking access to](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) [t](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies)[he Guardian](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) on 25th December 2022. +* **Azerbaijan and Russia block each other’s news media.** In [early June 2022](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm), Azerbaijan started [blocking access to Russia’s state-run RIA Novosti](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&domain=ria.ru) media website. Since (at least) 7th June 2022, Russian ISPs have been [blocking](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-06-01&until=2023-02-25&time_grain=day&probe_cc=RU&axis_y=domain) access to Azerbaijani news media websites (`haqqin.az`, `minval.az`, `oxu.az`, `ru.oxu.az`, `ru.baku.ws`). These blocks remain ongoing. +* **Temporary blocking of TikTok amid border clashes with Armenia.** During the September 2022 border clashes, both [Azerbaijan and Armenia blocked access to TikTok](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/). While the TikTok block was [lifted](https://explorer.ooni.org/search?since=2022-09-01&until=2022-09-30&failure=false&domain=www.tiktok.com&probe_cc=AM) in Armenia by 21st September 2022 (only lasting about a week), the [TikTok block](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) remained in place in Azerbaijan until November 2022 (lasting about 2 months). +* **Blocking of circumvention tool sites.** Azerbaijan continues to [block access to numerous circumvention tool websites](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=ANON), potentially limiting the ability to circumvent internet censorship in Azerbaijan. However, most OONI measurements suggest that tested circumvention tools ([Tor](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=tor) and [Psiphon](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=psiphon)) appear to be reachable. +* **Variance of censorship across networks.** While most ISPs in Azerbaijan appear to adopt similar censorship techniques (as we continued to observe connection timeouts in most anomalous measurements across ASNs), different ISPs block access to different websites over time. ## Methods -Since 2012, OONI has developed [free and open source software](https://github.com/ooni/), called [OONI Probe](https://ooni.org/install/), designed to [measure various forms of internet censorship](https://ooni.org/nettest/). OONI Probe is run by volunteers in around 160 countries every month, and their test results are automatically [published by OONI as open data in real-time](https://ooni.org/data/). More than [a billion network measurements](https://explorer.ooni.org/) have been collected and published from 25 thousand networks in 241 countries and territories over the last decade. +Since 2012, OONI has developed [free and open source software](https://github.com/ooni/), called [OONI Probe](https://ooni.org/install/), designed to [measure various forms of internet censorship](https://ooni.org/nettest/). OONI Probe is run by volunteers in around 160 countries every month, and their test results are automatically [published by OONI as open data in real-time](https://ooni.org/data/). More than [a billion network measurements](https://explorer.ooni.org/) have been collected and published from 25 thousand networks in 241 countries and territories over the last decade. -To examine internet censorship in Azerbaijan, we analyzed [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between January 2022 to February 2023. The goal of our analysis was to identify which websites and apps presented signs of blocking (“anomalies”) in Azerbaijan during the analysis period, particularly in comparison to blocks previously identified as part of our [past research](https://ooni.org/post/2021-azerbaijan/).   +To examine internet censorship in Azerbaijan, we analyzed [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between January 2022 to February 2023. The goal of our analysis was to identify which websites and apps presented signs of blocking (“anomalies”) in Azerbaijan during the analysis period, particularly in comparison to blocks previously identified as part of our [past research](https://ooni.org/post/2021-azerbaijan/). -Specifically, we analyzed measurements collected from [OONI’s Web Connectivity test](https://ooni.org/nettest/web-connectivity/), which is designed to measure the blocking of websites (these websites are publicly hosted on the [Citizen Lab test list Github repository](https://github.com/citizenlab/test-lists/tree/master/lists)). This test measures the accessibility of websites by attempting to perform a [DNS lookup](https://ooni.org/support/glossary/#dns-lookup), TCP/IP connection, and [HTTP GET request](https://ooni.org/support/glossary/#http-request) from two vantage points: (1) the local vantage point of the user and (2) a control network (non-censored network). The results from both networks are automatically compared and if they match, the tested URL is annotated as “accessible” (if the testing succeeds from the control vantage point). If the results differ, the tested URL is annotated as “[anomalous](https://ooni.org/support/faq/#what-do-you-mean-by-anomalies)”, which may provide a signal of potential blocking. +Specifically, we analyzed measurements collected from [OONI’s Web Connectivity test](https://ooni.org/nettest/web-connectivity/), which is designed to measure the blocking of websites (these websites are publicly hosted on the [Citizen Lab test list Github repository](https://github.com/citizenlab/test-lists/tree/master/lists)). This test measures the accessibility of websites by attempting to perform a [DNS lookup](https://ooni.org/support/glossary/#dns-lookup), TCP/IP connection, and [HTTP GET request](https://ooni.org/support/glossary/#http-request) from two vantage points: (1) the local vantage point of the user and (2) a control network (non-censored network). The results from both networks are automatically compared and if they match, the tested URL is annotated as “accessible” (if the testing succeeds from the control vantage point). If the results differ, the tested URL is annotated as “[anomalous](https://ooni.org/support/faq/#what-do-you-mean-by-anomalies)”, which may provide a signal of potential blocking. Depending on why the anomaly emerges, the anomalous measurement is automatically annotated as a DNS, TCP/IP, HTTP diff, or HTTP failure anomaly. For example, if the DNS lookup resolves to an IP address which differs from that resolved from the control vantage point, the measurement is annotated as a “DNS anomaly”, which may be a sign of [DNS tampering](https://ooni.org/support/glossary/#dns-tampering). -However, [false positives](https://ooni.org/support/faq/#what-are-false-positives) can occur, which is why we look at anomalous measurements in aggregate in order to determine if a tested URL consistently presents a large volume of anomalous measurements (in comparison to successful measurements) on a tested network. If a tested URL presents a large volume of anomalies, it may provide a stronger signal of potential blocking. If the types of anomalies are consistent (for example, always presenting DNS anomalies on a tested network), they offer an even stronger signal of potential censorship (since they suggest the use of a specific censorship technique, such as [DNS hijacking](https://ooni.org/support/glossary/#dns-hijacking)). But beyond aggregating anomalous measurements, we also analyze the raw data pertaining to anomalous measurements in order to identify the specific errors that occurred as part of the testing, offering insight into how a tested URL is potentially blocked. +However, [false positives](https://ooni.org/support/faq/#what-are-false-positives) can occur, which is why we look at anomalous measurements in aggregate in order to determine if a tested URL consistently presents a large volume of anomalous measurements (in comparison to successful measurements) on a tested network. If a tested URL presents a large volume of anomalies, it may provide a stronger signal of potential blocking. If the types of anomalies are consistent (for example, always presenting DNS anomalies on a tested network), they offer an even stronger signal of potential censorship (since they suggest the use of a specific censorship technique, such as [DNS hijacking](https://ooni.org/support/glossary/#dns-hijacking)). But beyond aggregating anomalous measurements, we also analyze the raw data pertaining to anomalous measurements in order to identify the specific errors that occurred as part of the testing, offering insight into how a tested URL is potentially blocked. -Based on our current heuristics, we automatically confirm the blocking of websites when a [block page](https://ooni.org/support/glossary/#block-page) is served and we have added the fingerprint of that blockpage to our database. We also automatically confirm the blocking of websites based on DNS answers containing IP addresses that are known to be associated with implementing internet censorship. For other forms of censorship, we analyze OONI data in order to aggregate anomalous measurements and identify why and how those anomalies occur, offering insight into additional cases of potential blocking. +Based on our current heuristics, we automatically confirm the blocking of websites when a [block page](https://ooni.org/support/glossary/#block-page) is served and we have added the fingerprint of that blockpage to our database. We also automatically confirm the blocking of websites based on DNS answers containing IP addresses that are known to be associated with implementing internet censorship. For other forms of censorship, we analyze OONI data in order to aggregate anomalous measurements and identify why and how those anomalies occur, offering insight into additional cases of potential blocking. ### Acknowledgement of limitations The findings of this study present several limitations, including: -* **Date range of analysis.** The findings are limited to [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between January 2022 to February 2023. As a result, findings from measurements collected in different date ranges are excluded from this study. -* **Type of measurements.** As part of this study, we primarily focus on OONI [Web Connectivity](https://ooni.org/nettest/web-connectivity/) measurements which pertain to the testing of websites for censorship. This focus was selected due to [known website censorship](https://ooni.org/post/2021-azerbaijan/) in the country, while aggregate views of OONI measurements from [other OONI Probe experiments](https://ooni.org/nettest) did not present an important volume of anomalies that would have warranted more dedicated analysis. -* **Tested websites.** OONI Probe website testing in Azerbaijan is primarily limited to URLs included in 2 [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists): the [global list](https://github.com/citizenlab/test-lists/blob/master/lists/global.csv) (including internationally-relevant URLs) and the [Azerbaijan list](https://github.com/citizenlab/test-lists/blob/master/lists/az.csv) (only including URLs relevant to Azerbaijan). As these lists are tested by [OONI Probe](https://ooni.org/install) users and there are bandwidth constraints, they are generally limited to around 1,000 URLs. As a result, the lists may exclude other websites which are blocked in Azerbaijan, and the findings are limited to the testing of the URLs included in these lists. Given that the lists are community-curated, we acknowledge the bias in terms of which URLs are added to the lists. -* **Testing coverage of websites.** Not all URLs included in [test lists](https://github.com/citizenlab/test-lists/tree/master/lists) are measured equally across Azerbaijan over time. Whether OONI data is available for a particular website depends on whether, on which networks, and when an [OONI Probe](https://ooni.org/install) user in Azerbaijan tested it. As a result, tested websites received different testing coverage throughout the analysis period, which impacts the findings.   -* **Tested ASNs.** The availability of OONI measurements depends on which networks [OONI Probe](https://ooni.org/install) users were connected to when performing tests. As a result, the measurement coverage varies across ASNs throughout the analysis period, impacting the findings. -* **Measurement volume.** [OONI measurement coverage from Azerbaijan](https://explorer.ooni.org/country/AZ) has been rather limited over the years (in comparison to other countries), limiting our findings and confidence in confirming censorship events. The availability of OONI data depends entirely on volunteers running [OONI Probe](https://ooni.org/install/) locally in Azerbaijan, which can be challenging in light of the political environment, [potential risks](https://ooni.org/about/risks/) involved with running OONI Probe, and the relatively limited digital rights community engagement opportunities in Azerbaijan. That said, we observe an [important spike in the overall measurement volume](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) from 14th September 2022 onwards (which has subsequently remained relatively stable), increasing our ability to detect censorship events thereafter. This is visible through the following chart. +* **Date range of analysis.** The findings are limited to [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) between January 2022 to February 2023. As a result, findings from measurements collected in different date ranges are excluded from this study. +* **Type of measurements.** As part of this study, we primarily focus on OONI [Web Connectivity](https://ooni.org/nettest/web-connectivity/) measurements which pertain to the testing of websites for censorship. This focus was selected due to [known website censorship](https://ooni.org/post/2021-azerbaijan/) in the country, while aggregate views of OONI measurements from [other OONI Probe experiments](https://ooni.org/nettest) did not present an important volume of anomalies that would have warranted more dedicated analysis. +* **Tested websites.** OONI Probe website testing in Azerbaijan is primarily limited to URLs included in 2 [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists): the [global list](https://github.com/citizenlab/test-lists/blob/master/lists/global.csv) (including internationally-relevant URLs) and the [Azerbaijan list](https://github.com/citizenlab/test-lists/blob/master/lists/az.csv) (only including URLs relevant to Azerbaijan). As these lists are tested by [OONI Probe](https://ooni.org/install) users and there are bandwidth constraints, they are generally limited to around 1,000 URLs. As a result, the lists may exclude other websites which are blocked in Azerbaijan, and the findings are limited to the testing of the URLs included in these lists. Given that the lists are community-curated, we acknowledge the bias in terms of which URLs are added to the lists. +* **Testing coverage of websites.** Not all URLs included in [test lists](https://github.com/citizenlab/test-lists/tree/master/lists) are measured equally across Azerbaijan over time. Whether OONI data is available for a particular website depends on whether, on which networks, and when an [OONI Probe](https://ooni.org/install) user in Azerbaijan tested it. As a result, tested websites received different testing coverage throughout the analysis period, which impacts the findings. +* **Tested ASNs.** The availability of OONI measurements depends on which networks [OONI Probe](https://ooni.org/install) users were connected to when performing tests. As a result, the measurement coverage varies across ASNs throughout the analysis period, impacting the findings. +* **Measurement volume.** [OONI measurement coverage from Azerbaijan](https://explorer.ooni.org/country/AZ) has been rather limited over the years (in comparison to other countries), limiting our findings and confidence in confirming censorship events. The availability of OONI data depends entirely on volunteers running [OONI Probe](https://ooni.org/install/) locally in Azerbaijan, which can be challenging in light of the political environment, [potential risks](https://ooni.org/about/risks/) involved with running OONI Probe, and the relatively limited digital rights community engagement opportunities in Azerbaijan. That said, we observe an [important spike in the overall measurement volume](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) from 14th September 2022 onwards (which has subsequently remained relatively stable), increasing our ability to detect censorship events thereafter. This is visible through the following chart. ![](images/image7.png) -**Chart:** OONI Probe Web Connectivity testing in Azerbaijan between 1st January 2022 to 23rd February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity)). +**Chart:** OONI Probe Web Connectivity testing in Azerbaijan between 1st January 2022 to 23rd February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity)). ## Background -In its latest [Freedom on the Net report](https://freedomhouse.org/country/azerbaijan/freedom-net/2022) by Freedom House, Azerbaijan was ranked “not free” among 70 countries assessed for the report. Internet freedoms have been on an overall decline in the country for a number of years, however, the situation escalated during the 2020 war between Armenia and Azerbaijan. Since then, while blocking of social media platforms, and internet throttling have reduced, the state continues to implement various forms of digital surveillance combined with offline measures targeting civil society representatives on a regular basis. +In its latest [Freedom on the Net report](https://freedomhouse.org/country/azerbaijan/freedom-net/2022) by Freedom House, Azerbaijan was ranked “not free” among 70 countries assessed for the report. Internet freedoms have been on an overall decline in the country for a number of years, however, the situation escalated during the 2020 war between Armenia and Azerbaijan. Since then, while blocking of social media platforms, and internet throttling have reduced, the state continues to implement various forms of digital surveillance combined with offline measures targeting civil society representatives on a regular basis. -News websites critical of the authorities remain blocked, while the [Law on Media](https://www.az-netwatch.org/news/azerbaijans-media-registry-leaves-media-platforms-in-limbo/) enacted in February 2022 is yet another measure imposed by the state to restrict not only media freedom, and freedom of speech but also access to independent and opposition news. +News websites critical of the authorities remain blocked, while the [Law on Media](https://www.az-netwatch.org/news/azerbaijans-media-registry-leaves-media-platforms-in-limbo/) enacted in February 2022 is yet another measure imposed by the state to restrict not only media freedom, and freedom of speech but also access to independent and opposition news. -Journalists across the country remain concerned about the new law. Most recently they have [joined](https://eurasianet.org/azerbaijani-journalists-fight-new-media-registry) in calls to abolish a new media registry introduced in the Media Law. +Journalists across the country remain concerned about the new law. Most recently they have [joined](https://eurasianet.org/azerbaijani-journalists-fight-new-media-registry) in calls to abolish a new media registry introduced in the Media Law. -Meanwhile, civic activists [continue to face](https://www.az-netwatch.org/news/a-year-in-review-from-online-attacks-to-overall-environment-of-internet-censorship-in-azerbaijan/) online targeting. Az-Net Watch in [its December 2022 report](https://www.az-netwatch.org/news/a-year-in-review-from-online-attacks-to-overall-environment-of-internet-censorship-in-azerbaijan/), shared some of the findings of such attacks, identifying persistent trends of phishing attacks, questioning by the law-enforcement bodies over criticism voiced online, interference with personal data and devices, hacking attempts and installed spyware programs. +Meanwhile, civic activists [continue to face](https://www.az-netwatch.org/news/a-year-in-review-from-online-attacks-to-overall-environment-of-internet-censorship-in-azerbaijan/) online targeting. Az-Net Watch in [its December 2022 report](https://www.az-netwatch.org/news/a-year-in-review-from-online-attacks-to-overall-environment-of-internet-censorship-in-azerbaijan/), shared some of the findings of such attacks, identifying persistent trends of phishing attacks, questioning by the law-enforcement bodies over criticism voiced online, interference with personal data and devices, hacking attempts and installed spyware programs. ## Findings -As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) (between January 2022 to February 2023), we found a number of independent news media websites and circumvention tool websites blocked throughout the analysis period. Many of these blocks have been in place [since at least 2020](https://ooni.org/post/2021-azerbaijan/). Amid border clashes in September 2022, access to [TikTok was blocked](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/) in both Azerbaijan and Armenia. We share further details in the sections below. +As part of our analysis of [OONI measurements collected from Azerbaijan](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity) (between January 2022 to February 2023), we found a number of independent news media websites and circumvention tool websites blocked throughout the analysis period. Many of these blocks have been in place [since at least 2020](https://ooni.org/post/2021-azerbaijan/). Amid border clashes in September 2022, access to [TikTok was blocked](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/) in both Azerbaijan and Armenia. We share further details in the sections below. ### Blocking of news media websites -Several news media websites (which have been [blocked over the last few years](https://ooni.org/post/2021-azerbaijan/#blocked-news-media-websites)) continue to be blocked in Azerbaijan, according to recent OONI data analysis. +Several news media websites (which have been [blocked over the last few years](https://ooni.org/post/2021-azerbaijan/#blocked-news-media-websites)) continue to be blocked in Azerbaijan, according to recent OONI data analysis. -The domains that presented the [largest volume of anomalies](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS) (and overall testing coverage), thereby presenting the strongest signals of blocking in our analysis period, include: +The domains that presented the [largest volume of anomalies](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS) (and overall testing coverage), thereby presenting the strongest signals of blocking in our analysis period, include: 1. `azerbaycansaati.tv` 2. `www.24saat.org` @@ -85,37 +85,37 @@ The domains that presented the [largest volume of anomalies](https://explorer.oo 9. `ria.ru` 10. `www.theguardian.com` -While most of the above domains have been [blocked in Azerbaijan over the last few years](https://ooni.org/post/2021-azerbaijan/#blocked-news-media-websites), the testing of `www.theguardian.com` only started to present anomalies (on several networks) in late December 2022 onwards (we discuss this case in more detail in the next section of this report). Meanwhile, Azerbaijan [reportedly started blocking access to RIA Novosti](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm) (Russian state-owned news media) in early June 2022 over the publication of slanderous materials against Azerbaijan. At the time, RIA Novosti’s website (\`ria.ru\`) was only included in other countries' [test lists](https://github.com/citizenlab/test-lists/tree/master/lists); we therefore [added](https://github.com/citizenlab/test-lists/pull/1000) it to the [Global test list](https://github.com/citizenlab/test-lists/blob/master/lists/global.csv) to ensure that it gets tested by [OONI Probe](https://ooni.org/install) users in Azerbaijan and around the world. This is why [OONI measurement coverage](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=ria.ru) of \`ria.ru\` in Azerbaijan only begins in June 2022.   +While most of the above domains have been [blocked in Azerbaijan over the last few years](https://ooni.org/post/2021-azerbaijan/#blocked-news-media-websites), the testing of `www.theguardian.com` only started to present anomalies (on several networks) in late December 2022 onwards (we discuss this case in more detail in the next section of this report). Meanwhile, Azerbaijan [reportedly started blocking access to RIA Novosti](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm) (Russian state-owned news media) in early June 2022 over the publication of slanderous materials against Azerbaijan. At the time, RIA Novosti’s website (\`ria.ru\`) was only included in other countries' [test lists](https://github.com/citizenlab/test-lists/tree/master/lists); we therefore [added](https://github.com/citizenlab/test-lists/pull/1000) it to the [Global test list](https://github.com/citizenlab/test-lists/blob/master/lists/global.csv) to ensure that it gets tested by [OONI Probe](https://ooni.org/install) users in Azerbaijan and around the world. This is why [OONI measurement coverage](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=ria.ru) of `ria.ru` in Azerbaijan only begins in June 2022. -Independent Azerbaijani media websites (such as `azadliq.info` and `meydan.tv`) have [reportedly](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/) been blocked since early 2017 for “posing a threat” to Azerbaijan’s national security. The state prosecutor reportedly accused these websites of sharing content that promotes violence, hatred, extremism, violates privacy and constitutes slander. However, the blocking of these media outlets may have been [politically motivated](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/). Azadliq, for example, reported on the [business dealings](https://www.azadliq.org/a/intihar-sui-qesd/28097510.html) of Vice President Mehriban Aliyeva’s private foundation, while the website of Gunun Sesi (which has [reportedly](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) been blocked since August 2018) is [operated](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) by Parviz Hashimli, a former political prisoner. The blocking of the RFE/RFL website (`www.rferl.org`) also [reportedly](https://www.rferl.org/a/azerbaijan-rferl-service-website-court-orders-blocked/28482679.html) began in 2017 following an Azerbaijani court order. +Independent Azerbaijani media websites (such as `azadliq.info` and `meydan.tv`) have [reportedly](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/) been blocked since early 2017 for “posing a threat” to Azerbaijan’s national security. The state prosecutor reportedly accused these websites of sharing content that promotes violence, hatred, extremism, violates privacy and constitutes slander. However, the blocking of these media outlets may have been [politically motivated](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/). Azadliq, for example, reported on the [business dealings](https://www.azadliq.org/a/intihar-sui-qesd/28097510.html) of Vice President Mehriban Aliyeva’s private foundation, while the website of Gunun Sesi (which has [reportedly](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) been blocked since August 2018) is [operated](https://freedomhouse.org/country/azerbaijan/freedom-net/2020) by Parviz Hashimli, a former political prisoner. The blocking of the RFE/RFL website (`www.rferl.org`) also [reportedly](https://www.rferl.org/a/azerbaijan-rferl-service-website-court-orders-blocked/28482679.html) began in 2017 following an Azerbaijani court order. The following chart aggregates OONI measurement coverage (from [26 ASNs](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&category_code=NEWS)) for the news media domains that presented the strongest signals of blocking in Azerbaijan between January 2022 to February 2023. ![](images/image15.png) -**Chart:** Blocked news media websites in Azerbaijan between January 2022 to February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS)). +**Chart:** Blocked news media websites in Azerbaijan between January 2022 to February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS)). -Most OONI measurements from the testing of these media websites (excluding `www.theguardian.com`, which is discussed in the next section) presented anomalies throughout the testing period, providing a signal of blocking. In contrast, most of the other URLs tested from the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists) during this period were found [accessible](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false) (on the same networks in Azerbaijan). +Most OONI measurements from the testing of these media websites (excluding `www.theguardian.com`, which is discussed in the next section) presented anomalies throughout the testing period, providing a signal of blocking. In contrast, most of the other URLs tested from the [Citizen Lab test lists](https://github.com/citizenlab/test-lists/tree/master/lists) during this period were found [accessible](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false) (on the same networks in Azerbaijan). -What is evident from the above chart is that the vast majority of [anomalous measurements](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&only=anomalies) presented [connection timeouts](https://explorer.ooni.org/measurement/20230223T180939Z_webconnectivity_AZ_28787_n1_bXGe6rZUa89gjAQQ?input=http%3A%2F%2Fazerbaycansaati.tv%2F) (for sites hosted on both HTTP and HTTPS). A consistent failure type can offer a strong signal of blocking, since it suggests the use of specific censorship techniques. In this case, the connection timeouts may indicate the use of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi) technology for the implementation of blocks, given that the timeout only happens after the TLS ClientHello message. The fact that we observe the prevalence of connection timeouts across anomalous measurements aggregated across tested ASNs further suggests that most ISPs in Azerbaijan adopt similar censorship techniques. +What is evident from the above chart is that the vast majority of [anomalous measurements](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&only=anomalies) presented [connection timeouts](https://explorer.ooni.org/measurement/20230223T180939Z_webconnectivity_AZ_28787_n1_bXGe6rZUa89gjAQQ?input=http%3A%2F%2Fazerbaycansaati.tv%2F) (for sites hosted on both HTTP and HTTPS). A consistent failure type can offer a strong signal of blocking, since it suggests the use of specific censorship techniques. In this case, the connection timeouts may indicate the use of [Deep Packet Inspection (DPI)](https://ooni.org/support/glossary/#dpi) technology for the implementation of blocks, given that the timeout only happens after the TLS ClientHello message. The fact that we observe the prevalence of connection timeouts across anomalous measurements aggregated across tested ASNs further suggests that most ISPs in Azerbaijan adopt similar censorship techniques. This is further evident when viewing a per-ASN breakdown of the specific failures in the measurement of each of the blocked news media sites. The following chart, for example, presents the specific failures that occurred when `www.24saat.org` was tested on multiple ASNs in Azerbaijan during our analysis period. ![](images/image12.png) -**Chart:** Measurement results (`OK` and specific failures) from the testing of `www.24saat.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.24saat.org)). +**Chart:** Measurement results (`OK` and specific failures) from the testing of `www.24saat.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.24saat.org)). -From the above chart we can see that most failures (across most tested ASNs) involve timeout errors (annotated in red). But we also observe some successful measurements (annotated as `OK` in green) on some of those networks too. To understand why those measurements were successful, we provide a further breakdown by testing targets. The following chart only presents measurement results (including specific failures observed in anomalous measurements) for the testing of the HTTPS version of the site (`https://www.24saat.org/`) across networks.   +From the above chart we can see that most failures (across most tested ASNs) involve timeout errors (annotated in red). But we also observe some successful measurements (annotated as `OK` in green) on some of those networks too. To understand why those measurements were successful, we provide a further breakdown by testing targets. The following chart only presents measurement results (including specific failures observed in anomalous measurements) for the testing of the HTTPS version of the site (`https://www.24saat.org/`) across networks. ![](images/image17.png) -**Chart:** Measurement results (`OK` and specific failures) from the testing of `https://www.24saat.org/` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&input=https%3A%2F%2Fwww.24saat.org%2F&probe_cc=AZ)). +**Chart:** Measurement results (`OK` and specific failures) from the testing of `https://www.24saat.org/` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&input=https%3A%2F%2Fwww.24saat.org%2F&probe_cc=AZ)). As we can see, the testing of the HTTPS version of the site was entirely successful on some networks, and presented consistent signs of blocking (consistently presenting connection timeouts) on others. In contrast, the testing of the HTTP version of the site (shared in the next chart below) shows that it presented signs of blocking on all of these tested networks. ![](images/image16.png) -**Chart:** Measurement results (`OK` and specific failures) from the testing of `http://www.24saat.org/` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&input=http%3A%2F%2Fwww.24saat.org%2F&probe_cc=AZ)). +**Chart:** Measurement results (`OK` and specific failures) from the testing of `http://www.24saat.org/` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&input=http%3A%2F%2Fwww.24saat.org%2F&probe_cc=AZ)). The fact that we observe more prevalent blocking of the HTTP version of the site (even though it supports HTTPS) suggests that the block is not implemented properly on certain networks. @@ -123,69 +123,69 @@ Similarly, we observe the prevalence of connection timeouts in the testing of ma ![](images/image11.png) -**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `azerbaycansaati.tv` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=azerbaycansaati.tv)). +**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `azerbaycansaati.tv` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=azerbaycansaati.tv)). The fact that the testing of `azerbaycansaati.tv` was not successful on any of the networks which received the largest volume of testing coverage, and that it consistently presented connection timeouts on most networks, provides a strong signal of blocking. -Similarly, we observe that the testing of `www.azadliq.org`, `www.meydan.tv`, and `www.rferl.org` mostly presented connection timeouts on tested networks, as illustrated through the following charts.    +Similarly, we observe that the testing of `www.azadliq.org`, `www.meydan.tv`, and `www.rferl.org` mostly presented connection timeouts on tested networks, as illustrated through the following charts. ![](images/image10.png) -**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.azadliq.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.azadliq.org)). +**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.azadliq.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.azadliq.org)). ![](images/image18.png) -**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.meydan.tv` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.meydan.tv)). +**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.meydan.tv` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.meydan.tv)). ![](images/image9.png) -**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.rferl.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.rferl.org)). +**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.rferl.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.rferl.org)). -OONI data also suggests that the site (www.occrp.org) of the Organized Crime and Corruption Reporting Project (OCCRP) is [blocked](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.occrp.org) in Azerbaijan as well. The blocking of the OCCRP site [reportedly](https://www.meydan.tv/en/article/occrp-blocked-in-azerbaijan/) began in September 2017, following the publication of a major investigation (“[Azerbaijani Laundromat](https://www.occrp.org/en/azerbaijanilaundromat/)”) into corruption, bribery, and money laundering in which powerful figures were allegedly involved. The following chart provides a per-ASN breakdown of the [measurement results](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.occrp.org) from the testing of \`www.occrp.org\`, showing that (similarly to the aforementioned blocked media sites) it presented connection timeouts on most networks. +OONI data also suggests that the site (www.occrp.org) of the Organized Crime and Corruption Reporting Project (OCCRP) is [blocked](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.occrp.org) in Azerbaijan as well. The blocking of the OCCRP site [reportedly](https://www.meydan.tv/en/article/occrp-blocked-in-azerbaijan/) began in September 2017, following the publication of a major investigation (“[Azerbaijani Laundromat](https://www.occrp.org/en/azerbaijanilaundromat/)”) into corruption, bribery, and money laundering in which powerful figures were allegedly involved. The following chart provides a per-ASN breakdown of the [measurement results](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.occrp.org) from the testing of `www.occrp.org\`, showing that (similarly to the aforementioned blocked media sites) it presented connection timeouts on most networks. ![](images/image4.png) -**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.occrp.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.occrp.org)). +**Chart:** Measurement results (including specific failures for anomalous measurements) from the testing of `www.occrp.org` on multiple ASNs in Azerbaijan (source: [OONI data](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.occrp.org)). Overall, the prevalence of connection timeouts (observed for different websites measured on many different networks in Azerbaijan) provides a strong signal of blocking. These blocks appear to have been implemented several years ago, and remain ongoing. #### Blocking of the Guardian? -On 25th December 2022, the testing of `www.theguardian.com` started to present anomalies in Azerbaijan. The following chart aggregates OONI measurement coverage from the testing of `www.theguardian.com` on [10 ASNs](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.theguardian.com) in Azerbaijan between 1st October 2022 to 24th February 2023. +On 25th December 2022, the testing of `www.theguardian.com` started to present anomalies in Azerbaijan. The following chart aggregates OONI measurement coverage from the testing of `www.theguardian.com` on [10 ASNs](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.theguardian.com) in Azerbaijan between 1st October 2022 to 24th February 2023. ![](images/image3.png) -**Chart:** OONI Probe testing of the Guardian (`www.theguardian.com`) on 10 ASNs in Azerbaijan between 1st October 2022 to 24th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.theguardian.com)). +**Chart:** OONI Probe testing of the Guardian (`www.theguardian.com`) on 10 ASNs in Azerbaijan between 1st October 2022 to 24th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.theguardian.com)). -However, these anomalies were [not](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.theguardian.com) [observed on all tested networks](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.theguardian.com) in Azerbaijan. To better understand the [anomalies](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies), we provide a breakdown of the specific failures observed in anomalous measurements on networks that received the largest testing coverage (through the following chart). +However, these anomalies were [not](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.theguardian.com) [observed on all tested networks](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-10-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=web_connectivity&domain=www.theguardian.com) in Azerbaijan. To better understand the [anomalies](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies), we provide a breakdown of the specific failures observed in anomalous measurements on networks that received the largest testing coverage (through the following chart). ![](images/image6.png) -**Chart:** Measurement results (including specific failures for anomalous measurements) from ASNs in Azerbaijan where the testing of `www.theguardian.com` presented anomalies between November 2022 to February 2023 (source: [OONI data](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com)). +**Chart:** Measurement results (including specific failures for anomalous measurements) from ASNs in Azerbaijan where the testing of `www.theguardian.com` presented anomalies between November 2022 to February 2023 (source: [OONI data](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com)). -As is evident from the above chart, all [anomalous measurements](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) presented connection timeout errors (for both the HTTP and HTTPS versions of the site). The prevalence of connection timeouts, coupled with the fact that it is the same failure that we also observe in the blocking of other news media websites (discussed previously), provides a signal of blocking. +As is evident from the above chart, all [anomalous measurements](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) presented connection timeout errors (for both the HTTP and HTTPS versions of the site). The prevalence of connection timeouts, coupled with the fact that it is the same failure that we also observe in the blocking of other news media websites (discussed previously), provides a signal of blocking. -While the data suggests that access to `www.theguardian.com` may be blocked on a few networks in Azerbaijan (particularly on [AS29049](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&probe_asn=AS29049) where we observe the largest volume of anomalies, which is consistent from 25th December 2022 onwards), it’s worth noting that it’s accessible on other networks (such as [AS28787](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&probe_asn=AS28787), where almost all measurements are successful). As we didn’t come across any news articles reporting the recent blocking of the Guardian in Azerbaijan (which could have helped corroborate the anomalous data), and the overall measurement volume is quite low, our confidence in confirming this potential block is rather limited.   +While the data suggests that access to `www.theguardian.com` may be blocked on a few networks in Azerbaijan (particularly on [AS29049](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&probe_asn=AS29049) where we observe the largest volume of anomalies, which is consistent from 25th December 2022 onwards), it’s worth noting that it’s accessible on other networks (such as [AS28787](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&probe_asn=AS28787), where almost all measurements are successful). As we didn’t come across any news articles reporting the recent blocking of the Guardian in Azerbaijan (which could have helped corroborate the anomalous data), and the overall measurement volume is quite low, our confidence in confirming this potential block is rather limited. #### Azerbaijani news media websites blocked in Russia -Following the [blocking](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm) of Russia’s state-run [RIA Novosti media website](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&domain=ria.ru) in Azerbaijan in early June 2022, we were informed that certain Azerbaijani news media websites (`haqqin.az`, `minval.az`, `oxu.az`, `ru.oxu.az`, `ru.baku.ws`) were reportedly blocked in Russia. On 7th June 2022, we [added these URLs to the Russian test list](https://github.com/citizenlab/test-lists/pull/1000/files) so that these websites could get tested by [OONI Probe](https://ooni.org/install) users in Russia. +Following the [blocking](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm) of Russia’s state-run [RIA Novosti media website](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&domain=ria.ru) in Azerbaijan in early June 2022, we were informed that certain Azerbaijani news media websites (`haqqin.az`, `minval.az`, `oxu.az`, `ru.oxu.az`, `ru.baku.ws`) were reportedly blocked in Russia. On 7th June 2022, we [added these URLs to the Russian test list](https://github.com/citizenlab/test-lists/pull/1000/files) so that these websites could get tested by [OONI Probe](https://ooni.org/install) users in Russia. -[OONI data](https://explorer.ooni.org/chart/mat?probe_cc=RU&since=2022-06-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity) confirms that these Azerbaijani websites have been blocked in Russia since at least 7th June 2022, and that these blocks remain ongoing (as illustrated below). +[OONI data](https://explorer.ooni.org/chart/mat?probe_cc=RU&since=2022-06-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity) confirms that these Azerbaijani websites have been blocked in Russia since at least 7th June 2022, and that these blocks remain ongoing (as illustrated below). ![](images/image20.png) -**Chart:** OONI data on the blocking of Azerbaijani news media websites in Russia between 7th June 2022 to 24th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-06-01&until=2023-02-25&time_grain=day&probe_cc=RU&axis_y=domain)). +**Chart:** OONI data on the blocking of Azerbaijani news media websites in Russia between 7th June 2022 to 24th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-06-01&until=2023-02-25&time_grain=day&probe_cc=RU&axis_y=domain)). -The above chart aggregates OONI measurement coverage from numerous networks in Russia, showing that at least 5 Azerbaijani news media domains were blocked in Russia between June 2022 to February 2023. Throughout this period, we observe that almost all measurements are anomalous, suggesting that access to these Azerbaijani media sites is blocked on most tested networks in Russia. Russian ISPs adopt a [variety of censorship techniques](https://ooni.org/post/2022-russia-blocks-amid-ru-ua-conflict/) and, on some networks, we are able to [automatically confirm the blocking](https://explorer.ooni.org/search?since=2022-06-07&until=2022-06-08&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&domain=haqqin.az&failure=false&only=confirmed) of these Azerbaijani news websites based on fingerprints (as IPs known to be associated with censorship are returned as part of DNS resolution). +The above chart aggregates OONI measurement coverage from numerous networks in Russia, showing that at least 5 Azerbaijani news media domains were blocked in Russia between June 2022 to February 2023. Throughout this period, we observe that almost all measurements are anomalous, suggesting that access to these Azerbaijani media sites is blocked on most tested networks in Russia. Russian ISPs adopt a [variety of censorship techniques](https://ooni.org/post/2022-russia-blocks-amid-ru-ua-conflict/) and, on some networks, we are able to [automatically confirm the blocking](https://explorer.ooni.org/search?since=2022-06-07&until=2022-06-08&probe_cc=RU&test_name=web_connectivity&category_code=NEWS&domain=haqqin.az&failure=false&only=confirmed) of these Azerbaijani news websites based on fingerprints (as IPs known to be associated with censorship are returned as part of DNS resolution). ### Temporary blocking of TikTok in Azerbaijan and Armenia amid border clashes -On 12th September 2022, fighting [erupted](https://www.bbc.com/news/world-europe-62828239) between Azerbaijani and Armenian troops along their border. The next day (13th September 2022), both [Azerbaijan and Armenia started blocking access to TikTok](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/) amid the border clashes. +On 12th September 2022, fighting [erupted](https://www.bbc.com/news/world-europe-62828239) between Azerbaijani and Armenian troops along their border. The next day (13th September 2022), both [Azerbaijan and Armenia started blocking access to TikTok](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/) amid the border clashes. -On 14th September 2022, Azerbaijan’s state security services [announced](https://dtx.gov.az/az/news/1715.html) the temporary suspension of TikTok. According to the statement, the content circulating on the social media platform was revealing military secrets and forming wrong opinions in society. As a result, the State Security Service decided to temporarily block access to the platform. Similarly, OONI data shows that the testing of TikTok also started to present signs of [blocking in Armenia](https://explorer.ooni.org/chart/mat?probe_cc=AM&test_name=web_connectivity&domain=tiktok.com&since=2022-08-16&until=2022-09-16&axis_x=measurement_start_day) from 13th September 2022 onwards. +On 14th September 2022, Azerbaijan’s state security services [announced](https://dtx.gov.az/az/news/1715.html) the temporary suspension of TikTok. According to the statement, the content circulating on the social media platform was revealing military secrets and forming wrong opinions in society. As a result, the State Security Service decided to temporarily block access to the platform. Similarly, OONI data shows that the testing of TikTok also started to present signs of [blocking in Armenia](https://explorer.ooni.org/chart/mat?probe_cc=AM&test_name=web_connectivity&domain=tiktok.com&since=2022-08-16&until=2022-09-16&axis_x=measurement_start_day) from 13th September 2022 onwards. -At the time, we published a [report](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/) documenting the block in both countries based on OONI data. Specifically, we found that [Azerbaijan](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/#blocking-of-tiktok-in-azerbaijan) blocked access to the main TikTok website (`www.tiktok.com`) and several endpoints essential to its functionality on at least 3 networks (AS29049, AS41997, AS31721). On all networks where we identified blocking, it seemed to be implemented by means of TLS level interference by dropping packets after noticing a disallowed server_name. We also observed some level of inconsistency, as ISPs in Azerbaijan blocked TikTok based on different lists of TikTok endpoint domains. +At the time, we published a [report](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/) documenting the block in both countries based on OONI data. Specifically, we found that [Azerbaijan](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/#blocking-of-tiktok-in-azerbaijan) blocked access to the main TikTok website (`www.tiktok.com`) and several endpoints essential to its functionality on at least 3 networks (AS29049, AS41997, AS31721). On all networks where we identified blocking, it seemed to be implemented by means of TLS level interference by dropping packets after noticing a disallowed server_name. We also observed some level of inconsistency, as ISPs in Azerbaijan blocked TikTok based on different lists of TikTok endpoint domains. In [Armenia](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/#blocking-of-tiktok-in-armenia), TikTok was blocked on at least two networks (AS44395, AS43733). Not all tested networks in Armenia implemented the TikTok block and censorship techniques varied from network to network. TikTok interference was observed at the DNS level by returning an NXDOMAIN error, an IP address associated with www.google.com, or a set of unrelated IP addresses which offer an invalid TLS certificate (it’s an expired certificate for `it.domain.name`). Several endpoints used by the TikTok app and website were also blocked on at least one network in Armenia. @@ -193,29 +193,29 @@ By comparing aggregate OONI measurements from the testing of TikTok’s website ![](images/image2.png) -**Chart:** OONI Probe testing of TikTok (`www.tiktok.com`) in Armenia and Azerbaijan between 1st September 2022 to 24th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?since=2022-09-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_cc&test_name=web_connectivity&domain=www.tiktok.com)). +**Chart:** OONI Probe testing of TikTok (`www.tiktok.com`) in Armenia and Azerbaijan between 1st September 2022 to 24th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?since=2022-09-01&until=2023-02-25&time_grain=day&axis_x=measurement_start_day&axis_y=probe_cc&test_name=web_connectivity&domain=www.tiktok.com)). -Specifically, we observe (through the above chart) that the TikTok block was [lifted](https://explorer.ooni.org/search?since=2022-09-01&until=2022-09-30&failure=false&domain=www.tiktok.com&probe_cc=AM) in Armenia by 21st September 2022 (only lasting about a week), while the [TikTok block](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) remained in place in Azerbaijan until November 2022 (lasting about 2 months). +Specifically, we observe (through the above chart) that the TikTok block was [lifted](https://explorer.ooni.org/search?since=2022-09-01&until=2022-09-30&failure=false&domain=www.tiktok.com&probe_cc=AM) in Armenia by 21st September 2022 (only lasting about a week), while the [TikTok block](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) remained in place in Azerbaijan until November 2022 (lasting about 2 months). -Similarly to the testing of `www.tiktok.com`, OONI data suggests that [Azerbaijan blocked numerous TikTok endpoints](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2022-11-30&time_grain=day&probe_cc=AZ&axis_y=domain) as well during the same period (between mid-September 2022 to mid-November 2022), as illustrated through the following two charts.   +Similarly to the testing of `www.tiktok.com`, OONI data suggests that [Azerbaijan blocked numerous TikTok endpoints](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2022-11-30&time_grain=day&probe_cc=AZ&axis_y=domain) as well during the same period (between mid-September 2022 to mid-November 2022), as illustrated through the following two charts.   ![](images/image19.png) -**Chart:** OONI Probe testing of TikTok endpoints in Azerbaijan between 1st September 2022 to 30th November 2022 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2022-11-30&time_grain=day&probe_cc=AZ&axis_y=domain)). +**Chart:** OONI Probe testing of TikTok endpoints in Azerbaijan between 1st September 2022 to 30th November 2022 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2022-11-30&time_grain=day&probe_cc=AZ&axis_y=domain)). ![](images/image13.png) -**Chart:** OONI Probe testing of TikTok endpoints in Azerbaijan between 1st September 2022 to 30th November 2022 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2022-11-30&time_grain=day&probe_cc=AZ&axis_y=domain)). +**Chart:** OONI Probe testing of TikTok endpoints in Azerbaijan between 1st September 2022 to 30th November 2022 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2022-11-30&time_grain=day&probe_cc=AZ&axis_y=domain)). -The fact that TikTok’s website (`www.tiktok.com`) and numerous TikTok endpoints presented anomalies on multiple tested networks in Azerbaijan during the same period provides a strong signal of blocking. +The fact that TikTok’s website (`www.tiktok.com`) and numerous TikTok endpoints presented anomalies on multiple tested networks in Azerbaijan during the same period provides a strong signal of blocking. -Between September to November 2022, the [anomalous measurements](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) from the testing of TikTok in Azerbaijan show that many of them time out when attempting to establish a TLS handshake to the target endpoint. While many such measurements are annotated (on [OONI Explorer](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies)) as presenting DNS inconsistency, we were able to exclude that by validating that the returned IPs are able to complete a TLS handshake with a valid certificate for the target domain name. Moreover, when issuing an HTTP request with the appropriate headers, the response payload is consistent with the response from an endpoint served to a user in Europe. By inspecting the response header  and server certificate, it seems to be an Akamai cache.   +Between September to November 2022, the [anomalous measurements](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) from the testing of TikTok in Azerbaijan show that many of them time out when attempting to establish a TLS handshake to the target endpoint. While many such measurements are annotated (on [OONI Explorer](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies)) as presenting DNS inconsistency, we were able to exclude that by validating that the returned IPs are able to complete a TLS handshake with a valid certificate for the target domain name. Moreover, when issuing an HTTP request with the appropriate headers, the response payload is consistent with the response from an endpoint served to a user in Europe. By inspecting the response header and server certificate, it seems to be an Akamai cache. ### Blocking of circumvention tool websites -Circumventing internet censorship in Azerbaijan can potentially be challenging, as [numerous circumvention tool websites have been blocked](https://ooni.org/post/2021-azerbaijan/#blocked-circumvention-tool-sites) in Azerbaijan over the past years. Our latest analysis shows that these blocks remain ongoing. +Circumventing internet censorship in Azerbaijan can potentially be challenging, as [numerous circumvention tool websites have been blocked](https://ooni.org/post/2021-azerbaijan/#blocked-circumvention-tool-sites) in Azerbaijan over the past years. Our latest analysis shows that these blocks remain ongoing. -The following chart aggregates [OONI measurement coverage](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-01-01&until=2023-02-24&time_grain=day&probe_cc=AZ&axis_y=domain&category_code=ANON) for popular circumvention tool websites that presented a large volume of anomalies throughout the testing period in Azerbaijan. +The following chart aggregates [OONI measurement coverage](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-01-01&until=2023-02-24&time_grain=day&probe_cc=AZ&axis_y=domain&category_code=ANON) for popular circumvention tool websites that presented a large volume of anomalies throughout the testing period in Azerbaijan. ![](images/image14.png) @@ -223,44 +223,44 @@ The following chart aggregates [OONI measurement coverage](https://explorer.ooni Multiple circumvention tool websites (including `www.torproject.org`, `psiphon.ca`, and `www.tunnelbear.com`) continue to show signs of blocking on some networks in Azerbaijan, as illustrated through the above chart. Specifically, OONI data shows that when these sites were tested, we frequently observed the [timing out of the session after the ClientHello during the TLS handshake](https://explorer.ooni.org/measurement/20221119T180719Z_webconnectivity_AZ_29049_n1_qOeKK5mXoM5sjD3e?input=https%3A%2F%2Fnordvpn.com%2F). It’s worth noting though that these sites are not blocked on all networks in Azerbaijan, as some measurements were successful. -While many popular circumvention tool sites appear to be blocked in Azerbaijan, it’s worth highlighting that certain circumvention tools may work in the country. [OONI Probe](https://ooni.org/install) also includes tests designed to measure the reachability of the [Psiphon](https://ooni.org/nettest/psiphon/), [Tor](https://ooni.org/nettest/tor/) and [Tor Snowflake](https://ooni.org/nettest/tor-snowflake/) circumvention tools. +While many popular circumvention tool sites appear to be blocked in Azerbaijan, it’s worth highlighting that certain circumvention tools may work in the country. [OONI Probe](https://ooni.org/install) also includes tests designed to measure the reachability of the [Psiphon](https://ooni.org/nettest/psiphon/), [Tor](https://ooni.org/nettest/tor/) and [Tor Snowflake](https://ooni.org/nettest/tor-snowflake/) circumvention tools. -Between January 2022 to February 2023, [OONI Probe testing of Psiphon](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=psiphon) shows that it was reachable on tested networks in Azerbaijan, as illustrated below. +Between January 2022 to February 2023, [OONI Probe testing of Psiphon](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=psiphon) shows that it was reachable on tested networks in Azerbaijan, as illustrated below. ![](images/image5.png) -**Chart:** OONI Probe testing of Psiphon on 29 ASNs in Azerbaijan between 1st January 2022 to 25th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=psiphon)). +**Chart:** OONI Probe testing of Psiphon on 29 ASNs in Azerbaijan between 1st January 2022 to 25th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=psiphon)). -Most measurements (collected from [29 ASNs](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=psiphon) during the testing period) show that it was possible to bootstrap Psiphon and use it to fetch a webpage, suggesting that the [circumvention tool](https://psiphon.ca/en/download.html?psiphonca) worked on tested networks in Azerbaijan. +Most measurements (collected from [29 ASNs](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=psiphon) during the testing period) show that it was possible to bootstrap Psiphon and use it to fetch a webpage, suggesting that the [circumvention tool](https://psiphon.ca/en/download.html?psiphonca) worked on tested networks in Azerbaijan. -While the testing of [Tor](https://www.torproject.org/) presented anomalies throughout the testing period, [most measurements](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=tor) from September 2022 onwards (when we also observe a spike in the overall measurement volume) suggest that [Tor was reachable](https://explorer.ooni.org/search?since=2022-09-14&until=2023-02-25&probe_cc=AZ&test_name=tor&failure=true) on most tested networks in Azerbaijan. +While the testing of [Tor](https://www.torproject.org/) presented anomalies throughout the testing period, [most measurements](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=tor) from September 2022 onwards (when we also observe a spike in the overall measurement volume) suggest that [Tor was reachable](https://explorer.ooni.org/search?since=2022-09-14&until=2023-02-25&probe_cc=AZ&test_name=tor&failure=true) on most tested networks in Azerbaijan. ![](images/image8.png) -**Chart:** OONI Probe testing of Tor on 28 ASNs in Azerbaijan between 1st January 2022 to 25th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=tor)). +**Chart:** OONI Probe testing of Tor on 28 ASNs in Azerbaijan between 1st January 2022 to 25th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-26&time_grain=day&axis_x=measurement_start_day&test_name=tor)). -In Tor anomalous measurements, we see that attempted connections to Tor directory authorities failed on some networks. However, we also see that most connections to both Tor directory authorities and [Tor bridges](https://bridges.torproject.org/) were [successful](https://explorer.ooni.org/measurement/20210415T185710Z_tor_AZ_39232_n1_tfLLJx1UKF2l9v4H) (on many networks), indicating that it may have been possible to use [Tor](https://www.torproject.org/download/) in Azerbaijan. +In Tor anomalous measurements, we see that attempted connections to Tor directory authorities failed on some networks. However, we also see that most connections to both Tor directory authorities and [Tor bridges](https://bridges.torproject.org/) were [successful](https://explorer.ooni.org/measurement/20210415T185710Z_tor_AZ_39232_n1_tfLLJx1UKF2l9v4H) (on many networks), indicating that it may have been possible to use [Tor](https://www.torproject.org/download/) in Azerbaijan. -Quite similarly, OONI data shows that it was possible to bootstrap [Tor Snowflake](https://snowflake.torproject.org/) on most networks in Azerbaijan throughout the testing period. +Quite similarly, OONI data shows that it was possible to bootstrap [Tor Snowflake](https://snowflake.torproject.org/) on most networks in Azerbaijan throughout the testing period. ![](images/image1.png) -**Chart:** OONI Probe testing of Tor Snowflake on 20 ASNs in Azerbaijan between 1st January 2022 to 25th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=torsf&axis_x=measurement_start_day&since=2022-01-01&until=2023-02-26&time_grain=day&probe_cc=AZ)). +**Chart:** OONI Probe testing of Tor Snowflake on 20 ASNs in Azerbaijan between 1st January 2022 to 25th February 2023 (source: [OONI data](https://explorer.ooni.org/chart/mat?test_name=torsf&axis_x=measurement_start_day&since=2022-01-01&until=2023-02-26&time_grain=day&probe_cc=AZ)). -This suggests that while the [Psiphon](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&probe_cc=AZ&test_name=web_connectivity&failure=true&domain=psiphon.ca&only=anomalies) and [Tor Project](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&probe_cc=AZ&test_name=web_connectivity&failure=true&domain=www.torproject.org&only=anomalies) websites appear to be blocked on some networks in Azerbaijan (as discussed previously), their tools appear to work in the country (at least on tested networks). +This suggests that while the [Psiphon](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&probe_cc=AZ&test_name=web_connectivity&failure=true&domain=psiphon.ca&only=anomalies) and [Tor Project](https://explorer.ooni.org/search?since=2022-01-01&until=2023-02-25&probe_cc=AZ&test_name=web_connectivity&failure=true&domain=www.torproject.org&only=anomalies) websites appear to be blocked on some networks in Azerbaijan (as discussed previously), their tools appear to work in the country (at least on tested networks). ## Conclusion -Azerbaijan continues to limit press freedom by [blocking access to several independent news media websites](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS) – the blocking of which appears to be [politically motivated](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/). OONI data also suggests that some ISPs in Azerbaijan [may have started blocking access to the Guardian](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) on 25th December 2022, but the relatively limited measurement coverage and the seemingly absence of news articles reporting the block limit our confidence in confirming this. +Azerbaijan continues to limit press freedom by [blocking access to several independent news media websites](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=NEWS) – the blocking of which appears to be [politically motivated](https://advox.globalvoices.org/2017/05/19/azerbaijan-blocks-independent-media-and-actually-admits-it/). OONI data also suggests that some ISPs in Azerbaijan [may have started blocking access to the Guardian](https://explorer.ooni.org/search?since=2022-11-01&until=2023-02-25&failure=false&probe_cc=AZ&domain=www.theguardian.com&only=anomalies) on 25th December 2022, but the relatively limited measurement coverage and the seemingly absence of news articles reporting the block limit our confidence in confirming this. -In early June 2022, both Azerbaijan and Russia started blocking access to each other’s news media websites. In early June 2022, Azerbaijan started [blocking access to Russia’s state-run RIA Novosti](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&domain=ria.ru) media website over the [publication of slanderous materials](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm) against Azerbaijan. Since (at least) 7th June 2022, Russian ISPs have been [blocking](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-06-01&until=2023-02-25&time_grain=day&probe_cc=RU&axis_y=domain) access to Azerbaijani news media websites (`haqqin.az`, `minval.az`, `oxu.az`, `ru.oxu.az`, `ru.baku.ws`). +In early June 2022, both Azerbaijan and Russia started blocking access to each other’s news media websites. In early June 2022, Azerbaijan started [blocking access to Russia’s state-run RIA Novosti](https://explorer.ooni.org/search?until=2023-02-24&since=2022-01-01&probe_cc=AZ&test_name=web_connectivity&failure=false&category_code=NEWS&domain=ria.ru) media website over the [publication of slanderous materials](https://www.turan.az/ext/news/2022/6/free/politics_news/ru/5722.htm) against Azerbaijan. Since (at least) 7th June 2022, Russian ISPs have been [blocking](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-06-01&until=2023-02-25&time_grain=day&probe_cc=RU&axis_y=domain) access to Azerbaijani news media websites (`haqqin.az`, `minval.az`, `oxu.az`, `ru.oxu.az`, `ru.baku.ws`). -Amid the [September 2022 border clashes](https://www.bbc.com/news/world-europe-62828239), both [Azerbaijan and Armenia started blocking access to TikTok](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/). However, while the TikTok block was [lifted](https://explorer.ooni.org/search?since=2022-09-01&until=2022-09-30&failure=false&domain=www.tiktok.com&probe_cc=AM) in Armenia by 21st September 2022 (only lasting about a week), the [TikTok block](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) remained in place in Azerbaijan until November 2022 (lasting about 2 months). But this was not the first time that Azerbaijan blocked access to a social media platform during conflict. Azerbaijan previously [blocked access to social media platforms during the 2020 Nagorno-Karabakh war](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war). +Amid the [September 2022 border clashes](https://www.bbc.com/news/world-europe-62828239), both [Azerbaijan and Armenia started blocking access to TikTok](https://ooni.org/post/2022-azerbaijan-and-armenia-blocks-tiktok/). However, while the TikTok block was [lifted](https://explorer.ooni.org/search?since=2022-09-01&until=2022-09-30&failure=false&domain=www.tiktok.com&probe_cc=AM) in Armenia by 21st September 2022 (only lasting about a week), the [TikTok block](https://explorer.ooni.org/search?since=2022-09-01&until=2022-11-30&failure=false&domain=www.tiktok.com&probe_cc=AZ&only=anomalies) remained in place in Azerbaijan until November 2022 (lasting about 2 months). But this was not the first time that Azerbaijan blocked access to a social media platform during conflict. Azerbaijan previously [blocked access to social media platforms during the 2020 Nagorno-Karabakh war](https://ooni.org/post/2021-azerbaijan/#blocking-of-social-media-amid-2020-nagorno-karabakh-war). -Both social media blocks during times of conflict and long-term news media blocks (as those seen in Azerbaijan) generally indicate government attempts to control political narratives. While numerous [circumvention tool websites appear to be blocked](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=ANON) in Azerbaijan (potentially limiting the ability to circumvent blocks), it’s worth noting that several circumvention tools (such as [Tor](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=tor) and [Psiphon](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=psiphon)) may work in the country. +Both social media blocks during times of conflict and long-term news media blocks (as those seen in Azerbaijan) generally indicate government attempts to control political narratives. While numerous [circumvention tool websites appear to be blocked](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&axis_y=domain&test_name=web_connectivity&category_code=ANON) in Azerbaijan (potentially limiting the ability to circumvent blocks), it’s worth noting that several circumvention tools (such as [Tor](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=tor) and [Psiphon](https://explorer.ooni.org/chart/mat?probe_cc=AZ&since=2022-01-01&until=2023-02-24&time_grain=day&axis_x=measurement_start_day&test_name=psiphon)) may work in the country. -This study was carried out through the use of [open methodologies](https://github.com/ooni/spec/tree/master/nettests), [free and open source software](https://ooni.org/install/), and [open data](https://ooni.org/data/), enabling independent third-party verification of our research findings. We encourage researchers to expand upon this study by [running OONI Probe](https://ooni.org/install/) and analyzing [OONI measurements from Azerbaijan](https://explorer.ooni.org/search?since=2023-01-26&until=2023-02-26&failure=false&probe_cc=AZ).   +This study was carried out through the use of [open methodologies](https://github.com/ooni/spec/tree/master/nettests), [free and open source software](https://ooni.org/install/), and [open data](https://ooni.org/data/), enabling independent third-party verification of our research findings. We encourage researchers to expand upon this study by [running OONI Probe](https://ooni.org/install/) and analyzing [OONI measurements from Azerbaijan](https://explorer.ooni.org/search?since=2023-01-26&until=2023-02-26&failure=false&probe_cc=AZ). ## Acknowledgements -We thank [OONI Probe](https://ooni.org/install) users in Azerbaijan for contributing measurements, supporting this study. +We thank [OONI Probe](https://ooni.org/install) users in Azerbaijan for contributing measurements, supporting this study. diff --git a/content/post/2023-brazil-telegram/index.md b/content/post/2023-brazil-telegram/index.md index f3df73d99..dbd7e36b0 100644 --- a/content/post/2023-brazil-telegram/index.md +++ b/content/post/2023-brazil-telegram/index.md @@ -3,75 +3,75 @@ title: "Brazil: OONI data on the blocking of Telegram" description: "This report shares OONI data on the blocking of Telegram in Brazil." author: "Arturo Filastò (OONI), Maria Xynou (OONI), Simone Basso (OONI), Felipe Octaviano Delgado Busnello (Independent researcher)" date: "2023-04-28" -tags: ["country-br", "censorship", "brazil", "telegram"] +tags: ["country-br", "censorship", "brazil", "telegram", "theme-im"] categories: ["report"] --- Encontre a versão em português do relatório localizada pela [Coding Rights](https://codingrights.org/) aqui: https://codingrights.org/library-item/brasil-dados-da-ooni-sobre-o-bloqueio-do-telegram/ -Two days ago, on 26th April 2023, Brazil started [blocking](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) access to Telegram. +Two days ago, on 26th April 2023, Brazil started [blocking](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) access to Telegram. -On the same day, a federal judge in Brazil [reportedly](https://www.dw.com/en/brazil-court-orders-temporary-suspension-of-telegram-over-neo-nazi-probe/a-65446224) ordered the temporary suspension of Telegram in response to the messaging service’s alleged failure to share all information on neo-Nazi chat groups requested by the police. This information was [reportedly](https://www.nytimes.com/2023/04/26/briefing/brazil-telegram-ban.html) requested as part of a school attacks inquiry, as Brazilian authorities investigate neo-Nazi groups that are believed to have used Telegram to incite school attacks. However, Telegram’s CEO [claims](https://www.washingtonpost.com/business/2023/04/27/telegram-brazil-block/2ec1c1de-e554-11ed-9696-8e874fd710b8_story.html) ([on his Telegram channel](https://web.telegram.org/z/#-1006503122)) that it’s impossible for them to comply with the order, as the judge requested data that is not feasible for them to obtain. +On the same day, a federal judge in Brazil [reportedly](https://www.dw.com/en/brazil-court-orders-temporary-suspension-of-telegram-over-neo-nazi-probe/a-65446224) ordered the temporary suspension of Telegram in response to the messaging service’s alleged failure to share all information on neo-Nazi chat groups requested by the police. This information was [reportedly](https://www.nytimes.com/2023/04/26/briefing/brazil-telegram-ban.html) requested as part of a school attacks inquiry, as Brazilian authorities investigate neo-Nazi groups that are believed to have used Telegram to incite school attacks. However, Telegram’s CEO [claims](https://www.washingtonpost.com/business/2023/04/27/telegram-brazil-block/2ec1c1de-e554-11ed-9696-8e874fd710b8_story.html) ([on his Telegram channel](https://web.telegram.org/z/#-1006503122)) that it’s impossible for them to comply with the order, as the judge requested data that is not feasible for them to obtain. -As of 26th April 2023, OONI data [shows](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) that some ISPs in Brazil started blocking access to Telegram. In this report, we share [OONI data collected from Brazil](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram), documenting the block. +As of 26th April 2023, OONI data [shows](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) that some ISPs in Brazil started blocking access to Telegram. In this report, we share [OONI data collected from Brazil](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram), documenting the block. {{}} ## Background -On 19th April 2023, a federal judge in Brazil issued a [decision](https://www.conjur.com.br/dl/telegram-decisao-suspensao.pdf) ordering Telegram to provide the court with the information requested by the Federal Police and the Public Prosecutors' Office. The information requested included the names, official ID numbers, profile pictures, profile statuses, email addresses, physical addresses, bank data, credit card information, linked devices (including IMEI), trusted telephone numbers for two-factor authentication, and account creation logs (including IP addresses, logical port, date, and time) of all users linked to the "Movimento Anti-Semita Brasileiro" channel and a neo-Nazi chat group named "Frente Anti-Semita". In the case of failure to comply with the order, the judge stipulated a daily fine of R$100.000 (roughly 20.000 USD) to the platform, or 5% of its annual gross income in Brazil, whichever is lower. +On 19th April 2023, a federal judge in Brazil issued a [decision](https://www.conjur.com.br/dl/telegram-decisao-suspensao.pdf) ordering Telegram to provide the court with the information requested by the Federal Police and the Public Prosecutors' Office. The information requested included the names, official ID numbers, profile pictures, profile statuses, email addresses, physical addresses, bank data, credit card information, linked devices (including IMEI), trusted telephone numbers for two-factor authentication, and account creation logs (including IP addresses, logical port, date, and time) of all users linked to the "Movimento Anti-Semita Brasileiro" channel and a neo-Nazi chat group named "Frente Anti-Semita". In the case of failure to comply with the order, the judge stipulated a daily fine of R$100.000 (roughly 20.000 USD) to the platform, or 5% of its annual gross income in Brazil, whichever is lower. -Telegram [responded](https://www.conjur.com.br/2023-abr-26/juiz-es-determina-suspensao-telegram-todo-brasil) by arguing that, as the groups ended their activities a few months prior to the order, they would be unable to provide the court with all the information requested. The platform was able to identify only the user number, the telephone number, and the creation date of the channel, as it had already been requested by the Federal Police in another case, while it was still active. As for the group, Telegram stated it would not be possible to recover any personal information unless the telephone number of the administrator was provided to them. +Telegram [responded](https://www.conjur.com.br/2023-abr-26/juiz-es-determina-suspensao-telegram-todo-brasil) by arguing that, as the groups ended their activities a few months prior to the order, they would be unable to provide the court with all the information requested. The platform was able to identify only the user number, the telephone number, and the creation date of the channel, as it had already been requested by the Federal Police in another case, while it was still active. As for the group, Telegram stated it would not be possible to recover any personal information unless the telephone number of the administrator was provided to them. -The Federal Police, however, stated that the information was not sufficient. The judge sided with the public institutions, remarked that the compliance to the order has been precarious, and stipulated that the information provided by Telegram did not meet the order. On 26th April 2023, the judge [ordered](https://www.conjur.com.br/dl/telegram-decisao-suspensao.pdf) the temporary cessation of Telegram's service in Brazil. This decision was forwarded to several Brazilian Internet Service Providers (ISPs). +The Federal Police, however, stated that the information was not sufficient. The judge sided with the public institutions, remarked that the compliance to the order has been precarious, and stipulated that the information provided by Telegram did not meet the order. On 26th April 2023, the judge [ordered](https://www.conjur.com.br/dl/telegram-decisao-suspensao.pdf) the temporary cessation of Telegram's service in Brazil. This decision was forwarded to several Brazilian Internet Service Providers (ISPs). The reasoning for the decision is that application providers on the internet, according to the [Brazilian Civil Rights Framework for the Internet law](http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2014/lei/l12965.htm), must store all data for a period of six months. Although Telegram alleged the group and the channel had already been deleted for more time than they are required to keep the data, according to the judge, Telegram has not proven such time had elapsed, and thus did not prove to have ridden itself of the legal obligation to keep the data. -Starting from 26th April 2023, several ISPs in Brazil started to [block](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) access to Telegram. Community members from Brazil shared OONI data on the block with their networks. Gustavo Gus from the Tor Project, for example, [reported](https://twitter.com/0xggus/status/1651738674365620225) the blocking of Telegram in Brazil by sharing relevant OONI data on social media. +Starting from 26th April 2023, several ISPs in Brazil started to [block](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) access to Telegram. Community members from Brazil shared OONI data on the block with their networks. Gustavo Gus from the Tor Project, for example, [reported](https://twitter.com/0xggus/status/1651738674365620225) the blocking of Telegram in Brazil by sharing relevant OONI data on social media. - {{Brazil Telegram}} +{{Brazil Telegram}} -As part of this report, we analyze [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) collected from Brazil to document the blocking of Telegram. +As part of this report, we analyze [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) collected from Brazil to document the blocking of Telegram. ## Methods -Founded in 2012, the [Open Observatory of Network Interference (OONI)](https://ooni.org/) is a free software project that aims to document internet censorship around the world. To this end, we build free software apps ([OONI Probe](https://ooni.org/install/)) which include [experiments](https://ooni.org/nettest/) designed to measure various forms of internet censorship, including the blocking of [Telegram](https://ooni.org/nettest/telegram). These experiments are run by [OONI Probe](https://ooni.org/install/) users in around 160 countries (including Brazil) every month, testing their networks to detect the blocking of websites and apps. To increase transparency of internet censorship, we publish OONI Probe test results (“measurements”) from around the world as [open data](https://ooni.org/data/) in real-time. +Founded in 2012, the [Open Observatory of Network Interference (OONI)](https://ooni.org/) is a free software project that aims to document internet censorship around the world. To this end, we build free software apps ([OONI Probe](https://ooni.org/install/)) which include [experiments](https://ooni.org/nettest/) designed to measure various forms of internet censorship, including the blocking of [Telegram](https://ooni.org/nettest/telegram). These experiments are run by [OONI Probe](https://ooni.org/install/) users in around 160 countries (including Brazil) every month, testing their networks to detect the blocking of websites and apps. To increase transparency of internet censorship, we publish OONI Probe test results (“measurements”) from around the world as [open data](https://ooni.org/data/) in real-time. -To investigate the blocking of Telegram in Brazil, we analyzed recent measurements collected from the OONI Probe testing of Telegram in the country. The [OONI Probe Telegram experiment](https://ooni.org/nettest/telegram) is designed to measure the reachability of Telegram’s app and web version within a tested network. +To investigate the blocking of Telegram in Brazil, we analyzed recent measurements collected from the OONI Probe testing of Telegram in the country. The [OONI Probe Telegram experiment](https://ooni.org/nettest/telegram) is designed to measure the reachability of Telegram’s app and web version within a tested network. -More specifically, the test attempts to establish a TCP connection to Telegram’s access points (DCs) and perform an HTTP POST request, as well as an HTTPS GET request to Telegram’s web version (web.telegram.org) over the vantage point of the user. The test results are annotated as “anomalous” (indicating potential blocking) when connections to all access points [defined in the test specification](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md) fail. We acknowledge that this may potentially present a methodological limitation, as it’s possible that ISPs may only block a few endpoints, but not all – and in this case, the measurement result would still be annotated as “successful” (since it requires that connections to all tested endpoints fail in order for the result to be annotated as “anomalous”). We therefore aim to [review our heuristics](https://github.com/ooni/probe/issues/2466) to improve the automated classification of results. +More specifically, the test attempts to establish a TCP connection to Telegram’s access points (DCs) and perform an HTTP POST request, as well as an HTTPS GET request to Telegram’s web version (web.telegram.org) over the vantage point of the user. The test results are annotated as “anomalous” (indicating potential blocking) when connections to all access points [defined in the test specification](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md) fail. We acknowledge that this may potentially present a methodological limitation, as it’s possible that ISPs may only block a few endpoints, but not all – and in this case, the measurement result would still be annotated as “successful” (since it requires that connections to all tested endpoints fail in order for the result to be annotated as “anomalous”). We therefore aim to [review our heuristics](https://github.com/ooni/probe/issues/2466) to improve the automated classification of results. Based on this [methodology](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md), Telegram’s app is likely blocked if TCP connections to all the tested Telegram access points fail, and/or if HTTP POST requests to Telegram’s access points do not send back a response to OONI’s servers. Telegram’s web version (web.telegram.org) is likely blocked if the HTTPS GET requests to web.telegram.org do not send back a consistent response to OONI’s servers. -However, [false positives](https://ooni.org/support/faq#what-are-false-positives) can occur due to a number of reasons, such as due to transient network failures, or if Telegram makes changes to their infrastructure that affect how the [OONI Probe Telegram experiment](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md) runs. We therefore analyze OONI Telegram measurements in aggregate with the goal of evaluating whether the test presents a relatively large volume of anomalies (in comparison to successful measurements) from each tested network during the date range of interest (in this case, comparing Telegram measurements collected from 26th April 2023 onwards with past Telegram measurements collected from Brazil over the last month). This aggregation is performed for all OONI Telegram measurements collected from all tested networks in Brazil. +However, [false positives](https://ooni.org/support/faq#what-are-false-positives) can occur due to a number of reasons, such as due to transient network failures, or if Telegram makes changes to their infrastructure that affect how the [OONI Probe Telegram experiment](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md) runs. We therefore analyze OONI Telegram measurements in aggregate with the goal of evaluating whether the test presents a relatively large volume of anomalies (in comparison to successful measurements) from each tested network during the date range of interest (in this case, comparing Telegram measurements collected from 26th April 2023 onwards with past Telegram measurements collected from Brazil over the last month). This aggregation is performed for all OONI Telegram measurements collected from all tested networks in Brazil. -While a comparatively (with past measurements) large volume of anomalous measurements can be indicative of Telegram blocking, it is necessary to inspect the raw measurement data to rule out false positives, understand why the anomalies occurred, and characterize the potential block. For this (more advanced) analysis, we used our [OONI data analysis tool](https://github.com/ooni/data) which enables us to more effectively analyze the raw measurement data, aggregate based on failure types, and characterize the block. +While a comparatively (with past measurements) large volume of anomalous measurements can be indicative of Telegram blocking, it is necessary to inspect the raw measurement data to rule out false positives, understand why the anomalies occurred, and characterize the potential block. For this (more advanced) analysis, we used our [OONI data analysis tool](https://github.com/ooni/data) which enables us to more effectively analyze the raw measurement data, aggregate based on failure types, and characterize the block. ## Findings -Starting from 27th April 2023, [OONI Probe testing of Telegram in Brazil](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) started to present a spike in anomalous measurements, as illustrated in the following chart. +Starting from 27th April 2023, [OONI Probe testing of Telegram in Brazil](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) started to present a spike in anomalous measurements, as illustrated in the following chart. {{Brazil Telegram}} -**Chart:** OONI Probe testing of Telegram on 378 ASNs in Brazil between 29th March 2023 to 28th April 2023 (source: [OONI MAT](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram)). +**Chart:** OONI Probe testing of Telegram on 378 ASNs in Brazil between 29th March 2023 to 28th April 2023 (source: [OONI MAT](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram)). The above chart aggregates OONI Telegram measurement coverage from 378 ASNs in Brazil between 29th March 2023 to 28th April 2023. Successful Telegram measurements (indicating Telegram reachability) are annotated as “OK” (green), while measurements that present anomalies (signs of potential blocking) are annotated as “Anomaly” (orange). As is evident, almost all measurements over the past month were successful, and in the chart above, we only observe a noticeable increase in anomalous measurements from 27th April 2023 onwards (which matches the timing of the reported block). The comparatively (with past measurements) large volume of anomalous measurements (along with the timing of the anomalies and the fact that they persisted the next day) provides a signal of Telegram blocking in Brazil. However, we also observe that some measurements (on 27th and 28th April 2023) were successful, indicating that if Telegram is blocked, it may not be blocked on all tested networks. -A [per-ASN breakdown of the measurement coverage](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=telegram) suggests that Telegram is not blocked on all tested networks in Brazil. OONI data, for example, shows that Telegram was reachable when tested on [AS262374](https://ipinfo.io/AS262374) on 28th April 2023, as the relevant measurements [show](https://explorer.ooni.org/search?since=2023-04-28&until=2023-04-29&probe_cc=BR&test_name=telegram&failure=true&probe_asn=AS262374) that connections to the tested Telegram endpoints were successful. However, it is possible that blocking may have been implemented for Telegram endpoints not currently included in the [OONI Probe Telegram experiment](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md). Moreover, the fact that the OONI Probe Telegram experiment [requires all tested endpoints to fail](https://github.com/ooni/probe/issues/2466) in order to be annotated as “anomalous” means that it’s possible that in cases where only a few endpoints are blocked (instead of all), those measurements would be annotated as “successful”. We implemented this heuristic because, when we wrote the experiment, it was sufficient for a single DC endpoint to be accessible for the Telegram app to work as intended. As mentioned previously, we may need to [review our heuristics](https://github.com/ooni/probe/issues/2466). +A [per-ASN breakdown of the measurement coverage](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=telegram) suggests that Telegram is not blocked on all tested networks in Brazil. OONI data, for example, shows that Telegram was reachable when tested on [AS262374](https://ipinfo.io/AS262374) on 28th April 2023, as the relevant measurements [show](https://explorer.ooni.org/search?since=2023-04-28&until=2023-04-29&probe_cc=BR&test_name=telegram&failure=true&probe_asn=AS262374) that connections to the tested Telegram endpoints were successful. However, it is possible that blocking may have been implemented for Telegram endpoints not currently included in the [OONI Probe Telegram experiment](https://github.com/ooni/spec/blob/master/nettests/ts-020-telegram.md). Moreover, the fact that the OONI Probe Telegram experiment [requires all tested endpoints to fail](https://github.com/ooni/probe/issues/2466) in order to be annotated as “anomalous” means that it’s possible that in cases where only a few endpoints are blocked (instead of all), those measurements would be annotated as “successful”. We implemented this heuristic because, when we wrote the experiment, it was sufficient for a single DC endpoint to be accessible for the Telegram app to work as intended. As mentioned previously, we may need to [review our heuristics](https://github.com/ooni/probe/issues/2466). -The following chart shares the [ASNs](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=telegram) which presented the largest volume of anomalies (and, therefore, the strongest signals of blocking) from the OONI Probe testing of Telegram in Brazil. +The following chart shares the [ASNs](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=telegram) which presented the largest volume of anomalies (and, therefore, the strongest signals of blocking) from the OONI Probe testing of Telegram in Brazil. {{Brazil Telegram}} -**Chart:** ASNs in Brazil where the OONI Probe testing of Telegram presented the largest volume of anomalies between 29th March 2023 to 28th April 2023 (source: [OONI MAT](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=telegram)). +**Chart:** ASNs in Brazil where the OONI Probe testing of Telegram presented the largest volume of anomalies between 29th March 2023 to 28th April 2023 (source: [OONI MAT](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&axis_y=probe_asn&test_name=telegram)). -By performing a more detailed analysis using our [OONI data analysis tool](https://github.com/ooni/data) we are able to look at the Telegram reachability tests on a per-TCP-endpoint basis. In the chart below, we can see “TCP connect” observations. Each row represents a distinct TCP endpoint (an IP and port pair), while the columns are distinct network names (the same network name may be used by several different AS numbers). We limited this chart to some selected networks, focusing on the most popular ones, as well as those which had sufficient coverage during the measurement period and highlighted some differences in blocking of particular endpoints. +By performing a more detailed analysis using our [OONI data analysis tool](https://github.com/ooni/data) we are able to look at the Telegram reachability tests on a per-TCP-endpoint basis. In the chart below, we can see “TCP connect” observations. Each row represents a distinct TCP endpoint (an IP and port pair), while the columns are distinct network names (the same network name may be used by several different AS numbers). We limited this chart to some selected networks, focusing on the most popular ones, as well as those which had sufficient coverage during the measurement period and highlighted some differences in blocking of particular endpoints. {{Brazil Telegram}} -**Chart:** Blocking on selected Telegram endpoints across selected Brazilian ISPs. Each row in the chart shows the status of a specific endpoint and each column of a specific ISP (source: [ooni/data](https://github.com/ooni/data) tool). +**Chart:** Blocking on selected Telegram endpoints across selected Brazilian ISPs. Each row in the chart shows the status of a specific endpoint and each column of a specific ISP (source: [ooni/data](https://github.com/ooni/data) tool). What we can see from the above chart is that most networks block all endpoints starting from around 27th April 2023. In some cases, however, the provider did not block all the measured endpoints effectively, but only chose to block some of them. For example, on TIM S/A (AS26615), the four top endpoints are not blocked. Since we observe consistent IP blocking regardless of the port, it’s plausible to speculate that blocking depends on IP address rather than IP address and port. On Ligga Telecomunicações S.A. (AS14868), we observe different endpoints being accessible (from 7th to 10th April 2023 in the chart above). This leads us to believe that there is no information sharing amongst providers of the targets for the block and each provider chooses what to block autonomously. On Unifique Telecomunicações SA (AS28343), we observe no evidence of blocking at all. @@ -81,14 +81,14 @@ By looking at the blocking time for a particular endpoint across providers, we c {{Brazil Telegram}} -**Chart**: Blocking of a specific Telegram endpoint for distinct ISPs. From the chart, we can see how each ISP started the blocking at a different point in time (source: [ooni/data](https://github.com/ooni/data) tool). +**Chart**: Blocking of a specific Telegram endpoint for distinct ISPs. From the chart, we can see how each ISP started the blocking at a different point in time (source: [ooni/data](https://github.com/ooni/data) tool). As is evident through the above chart, TIM was the first provider to implement the block at around 23:00 UTC on 26th April 2023, followed by Telefônica Brasil an hour later, followed by Claro the following hour. Several hours later, at around 14:00 UTC on 27th April 2023, V tal (also known as Oi) started blocking Telegram, followed an hour later by Ligga. ## Conclusion -OONI data [shows](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) that some ISPs in Brazil immediately complied with the federal judge’s [decision](https://www.conjur.com.br/dl/telegram-decisao-suspensao.pdf) to suspend Telegram, as the blocking of Telegram started on 26th April 2023. However, the block was not implemented by all ISPs in Brazil, nor was it implemented in the same way. Some ISPs blocked access to all tested Telegram endpoints, while others only blocked some of them. This suggests lack of coordination between providers, and that each ISP implemented the block autonomously. +OONI data [shows](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram) that some ISPs in Brazil immediately complied with the federal judge’s [decision](https://www.conjur.com.br/dl/telegram-decisao-suspensao.pdf) to suspend Telegram, as the blocking of Telegram started on 26th April 2023. However, the block was not implemented by all ISPs in Brazil, nor was it implemented in the same way. Some ISPs blocked access to all tested Telegram endpoints, while others only blocked some of them. This suggests lack of coordination between providers, and that each ISP implemented the block autonomously. -This is not the first time that access to a popular messaging service is blocked in Brazil over noncompliance with authorities’ requests to hand over user data. WhatsApp, for example, was [temporarily blocked](https://ooni.org/post/brazil-whatsapp-block/) in Brazil (in December 2015 and May 2016) over noncompliance with such requests. +This is not the first time that access to a popular messaging service is blocked in Brazil over noncompliance with authorities’ requests to hand over user data. WhatsApp, for example, was [temporarily blocked](https://ooni.org/post/brazil-whatsapp-block/) in Brazil (in December 2015 and May 2016) over noncompliance with such requests. -You can continue to monitor the blocking of Telegram in Brazil through [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram), which is automatically published in real-time. We thank [OONI Probe](https://ooni.org/install/) users in Brazil who contributed measurements, supporting this study. +You can continue to monitor the blocking of Telegram in Brazil through [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=BR&since=2023-03-29&until=2023-04-29&time_grain=day&axis_x=measurement_start_day&test_name=telegram), which is automatically published in real-time. We thank [OONI Probe](https://ooni.org/install/) users in Brazil who contributed measurements, supporting this study. diff --git a/content/post/2023-ethiopia-blocks-social-media/index.md b/content/post/2023-ethiopia-blocks-social-media/index.md index 5caad2ff0..0656db40f 100644 --- a/content/post/2023-ethiopia-blocks-social-media/index.md +++ b/content/post/2023-ethiopia-blocks-social-media/index.md @@ -3,7 +3,7 @@ title: "Ethiopia: Ongoing blocking of social media" description: "This report shares OONI data on the blocking of social media platforms in Ethiopia." author: "Maria Xynou, Arturo Filastò" date: "2023-02-15" -tags: ["ethiopia", "country-et", "censorship"] +tags: ["ethiopia", "country-et", "censorship", "theme-social_media", "theme-im"] categories: ["report"] --- diff --git a/content/post/2023-jordan-blocks-grindr/index.md b/content/post/2023-jordan-blocks-grindr/index.md index dc13fa29e..e0e0057f3 100644 --- a/content/post/2023-jordan-blocks-grindr/index.md +++ b/content/post/2023-jordan-blocks-grindr/index.md @@ -3,7 +3,7 @@ title: "Grindr blocked in Jordan: Shrinking LGBTQ spaces" description: "This report shares OONI data on the blocking of Grindr in Jordan." author: "Maria Xynou (OONI), Arturo Filastò (OONI), My Kali Magazine" date: "2023-09-20" -tags: ["country-jo", "censorship", "jordan", "grindr"] +tags: ["country-jo", "censorship", "jordan", "grindr", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/2023-pakistan-blocks-wikipedia-and-dw/index.md b/content/post/2023-pakistan-blocks-wikipedia-and-dw/index.md index 5c79bcb07..ba2f8b4b8 100644 --- a/content/post/2023-pakistan-blocks-wikipedia-and-dw/index.md +++ b/content/post/2023-pakistan-blocks-wikipedia-and-dw/index.md @@ -3,7 +3,7 @@ title: "Pakistan: Blocking of Wikipedia and Deutsche Welle (DW)" description: "This report shares OONI data on the blocking of Wikipedia and DW in Pakistan." author: "Maria Xynou, Arturo Filastò" date: "2023-02-15" -tags: ["pakistan", "censorship", "country-pk", "wikipedia"] +tags: ["pakistan", "censorship", "country-pk", "wikipedia", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/2023-russia-a-year-after-the-conflict/index.md b/content/post/2023-russia-a-year-after-the-conflict/index.md index 19cfcd4f1..812a3aea5 100644 --- a/content/post/2023-russia-a-year-after-the-conflict/index.md +++ b/content/post/2023-russia-a-year-after-the-conflict/index.md @@ -3,7 +3,7 @@ title: "How Internet censorship changed in Russia during the 1st year of militar description: "This report shares OONI data on new blocks that emerged in Russia during the 1st year of military conflict in Ukraine." author: "Roskomsvoboda, OONI" date: "2023-02-24" -tags: ["country-ru", "censorship", "russia"] +tags: ["country-ru", "censorship", "russia", "theme-human_rights", "theme-news_media", "theme-social_media", "theme-circumvention"] categories: ["report"] --- @@ -26,7 +26,7 @@ New blocks that emerged over the last year include: * Blocking of **investigative journalism** ([Agentura.Ru](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-01-01&until=2023-02-21&time_grain=day&probe_cc=RU&domain=www.agentura.ru)) and **independent news media** sites ([The Bell](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2023-01-20&until=2023-02-21&time_grain=day&probe_cc=RU&domain=thebell.io)); * Blocking of [Instagram](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-01-01&until=2023-02-22&time_grain=day&probe_cc=RU&domain=www.instagram.com), [SoundCloud](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-09-01&until=2023-02-19&time_grain=day&probe_cc=RU&domain=soundcloud.com), and [Patreon](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-07-01&until=2023-02-19&time_grain=day&probe_cc=RU&domain=www.patreon.com). -Access to the Tor Project’s website was [temporarily unblocked](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-01-01&until=2022-12-31&time_grain=day&probe_cc=RU&domain=www.torproject.org) (between 15th to 28th July 2022), while several discontinued sites have been unblocked (possibly as part of a “blocklist cleanup”). +Access to the Tor Project’s website was [temporarily unblocked](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2022-01-01&until=2022-12-31&time_grain=day&probe_cc=RU&domain=www.torproject.org) (between 15th to 28th July 2022), while several discontinued sites have been unblocked (possibly as part of a “blocklist cleanup”). ## Background diff --git a/content/post/2023-russia-a-year-after-the-conflict/index.ru.md b/content/post/2023-russia-a-year-after-the-conflict/index.ru.md index f31bc909c..3afb08fa1 100644 --- a/content/post/2023-russia-a-year-after-the-conflict/index.ru.md +++ b/content/post/2023-russia-a-year-after-the-conflict/index.ru.md @@ -3,7 +3,7 @@ title: "Год военной цензуры в России" description: "Этот отчет анализирует данные OONI о новых блокировках, произошедших в России в течение первого года военного конфликта в Украине." author: "Roskomsvoboda, OONI" date: "2023-02-24" -tags: ["country-ru", "censorship", "russia"] +tags: ["country-ru", "censorship", "russia", "theme-human_rights", "theme-news_media", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2023-senegal-social-media-blocks/index.md b/content/post/2023-senegal-social-media-blocks/index.md index 46fae1600..08c7f36eb 100644 --- a/content/post/2023-senegal-social-media-blocks/index.md +++ b/content/post/2023-senegal-social-media-blocks/index.md @@ -3,7 +3,7 @@ title: "Senegal: Social media blocks and network outages amid political unrest" description: "This report documents the blocking of social media and network outages amid political unrest in Senegal." author: "Laura Schwartz-Henderson (Independent Consultant), David Belson (Cloudflare), Zach Rosson (Access Now), Felicia Anthonio (Access Now), Maria Xynou (OONI), Arturo Filastò (OONI)" date: "2023-08-01" -tags: ["country-sn"] +tags: ["country-sn", "theme-social_media", "theme-im"] categories: ["report"] --- @@ -23,7 +23,7 @@ OONI data collected from Senegal shows the blocking of [WhatsApp](https://explor The survey findings suggest that these internet disruptions had widespread impact. Senegalese survey respondents who experienced the outages and blocks on the ground described that they had an impact on Senegal’s economic activity and political stability, on their professional endeavors, their social and psychological health and well-being, and on the vulnerable or marginalized groups in Senegal. -## Background  +## Background Senegal is continuing a worrying [trend of democratic backsliding](https://preparepreventresist.org/senegal-2/) through the escalating [use of internet shutdowns](https://www.accessnow.org/press-release/keep-people-connected-senegal/) during protests. On 4th March 2021, following protests against the high-profile arrest of opposition leader Ousmane Sonko, Senegalese authorities [reportedly blocked](https://www.bloomberg.com/news/articles/2021-03-05/senegal-shuts-down-tv-stations-internet-disrupted-amid-protests#xj4y7vzkg) Facebook, YouTube, WhatsApp, and Telegram. Additionally, President Macky Sall has repeatedly [spoken out](https://www.theafricareport.com/99491/senegals-president-joins-the-chorus-against-twitter-and-facebook/) against social media, expressing desire to more tightly control online spaces and suppress dissent. These are not idle threats, considering a measurably [declining press freedom landscape](https://www.occrp.org/en/daily/17009-senegalese-journalist-charged-with-endangering-national-security) and now at least two sets of internet shutdowns since 2021. diff --git a/content/post/2023-throttling-kz-elections/index.md b/content/post/2023-throttling-kz-elections/index.md index c02912b5f..57e83cd4a 100644 --- a/content/post/2023-throttling-kz-elections/index.md +++ b/content/post/2023-throttling-kz-elections/index.md @@ -3,7 +3,7 @@ title: "Throttling of news media amid Kazakhstan’s 2022 presidential election" description: "A technical report highlighting the throttling of Radio Free Europe/Radio Liberty websites in Kazakhstan from September 2022 to April 2023, alongside temporary throttling of www.currenttime.tv during the 2022 presidential elections" author: "Arturo Filastò, Simone Basso" date: "2023-04-28" -tags: ["kazakhstan", "censorship", "country-kz", "rferl"] +tags: ["kazakhstan", "censorship", "country-kz", "rferl", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/2023-turkey-throttling-blocking-twitter/index.md b/content/post/2023-turkey-throttling-blocking-twitter/index.md index 7634fbec1..87fb15f7f 100644 --- a/content/post/2023-turkey-throttling-blocking-twitter/index.md +++ b/content/post/2023-turkey-throttling-blocking-twitter/index.md @@ -3,7 +3,7 @@ title: "Turkey: Throttling and DNS blocking of Twitter following deadly earthqua description: "This report shares OONI data on the throttling and DNS blocking of Twitter in Turkey." author: "Maria Xynou, Arturo Filastò" date: "2023-02-15" -tags: ["turkey", "country-tr", "censorship"] +tags: ["turkey", "country-tr", "censorship", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/2024-kazakhstan-report/index.md b/content/post/2024-kazakhstan-report/index.md index 372dd48a3..84d30eacd 100644 --- a/content/post/2024-kazakhstan-report/index.md +++ b/content/post/2024-kazakhstan-report/index.md @@ -3,7 +3,7 @@ title: "Kazakhstan: TLS MITM attacks and blocking of news media, human rights, a description: "This report documents TLS MITM attacks and the blocking of news media, human rights and circumvention tools websites in Kazakhstan based on OONI data." author: "Yelzhan Kabyshev (Head of Legal Practice, Eurasian Digital Foundation, Manager of the Human Rights Project, Internet Freedom Kazakhstan), Ruslan Daiyrbekov (Founder, Eurasian Digital Foundation), Vadim Melyakov (Analyst, Eurasian Digital Foundation, Specialist in International Relations), Igor Loskutov (Head of the Legal Department, InfoTech&Service LLP), Maria Xynou (OONI), Elizaveta Yachmeneva (OONI), Arturo Filastò (OONI), Mehul Gulati (OONI)" date: "2024-09-19" -tags: ["country-kz", "censorship", "kazakhstan"] +tags: ["country-kz", "censorship", "kazakhstan", "theme-news_media", "theme-human_rights", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2024-kazakhstan-report/index.ru.md b/content/post/2024-kazakhstan-report/index.ru.md index f3b0d0513..2b6553b51 100644 --- a/content/post/2024-kazakhstan-report/index.ru.md +++ b/content/post/2024-kazakhstan-report/index.ru.md @@ -3,7 +3,7 @@ title: "Казахстан: TLS MITM атаки и блокировка сайт description: "Этот отчет анализирует данные OONI о новых блокировках сайтов медиа, правозащитных проектов и средств по обходу блокировок, произошедших в Казахстане." author: "Yelzhan Kabyshev (Head of Legal Practice, Eurasian Digital Foundation, Manager of the Human Rights Project, Internet Freedom Kazakhstan), Ruslan Daiyrbekov (Founder, Eurasian Digital Foundation), Vadim Melyakov (Analyst, Eurasian Digital Foundation, Specialist in International Relations), Igor Loskutov (Head of the Legal Department, InfoTech&Service LLP), Maria Xynou (OONI), Elizaveta Yachmeneva (OONI), Arturo Filastò (OONI), Mehul Gulati (OONI)" date: "2024-09-19" -tags: ["country-kz", "censorship", "kazakhstan"] +tags: ["country-kz", "censorship", "kazakhstan", "theme-news_media", "theme-human_rights", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.md b/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.md index 4e04b8699..a48676629 100644 --- a/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.md +++ b/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.md @@ -3,7 +3,7 @@ title: "Tanzania: Surge in online LGBTIQ censorship and other targeted blocks" description: "This report documents the blocking of LGBTIQ websites and other targeted blocks in Tanzania based on OONI data." author: "Maria Xynou (OONI), Mehul Gulati (OONI), Tori Fransis" date: "2024-04-11" -tags: ["country-tz"] +tags: ["country-tz", "theme-human_rights", "theme-social_media", "theme-circumvention"] categories: ["report"] --- @@ -24,7 +24,7 @@ Specifically, [OONI data](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since= * LGBTIQ suicide prevention ([The Trevor Project](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2023-12-01&until=2024-01-31&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.thetrevorproject.org)) * **Blocking of online dating websites** (such as [Tinder](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2023-12-01&until=2024-01-31&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=tinder.com) and [OKCupid](https://explorer.ooni.org/chart/mat?test_name=web_connectivity&axis_x=measurement_start_day&since=2023-01-01&until=2024-01-31&time_grain=day&probe_cc=TZ&domain=www.okcupid.com)) -* **Blocking of specific websites that** **defend human rights through grants and petitions** ([Change.org](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.change.org), [Global Fund for Women](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.globalfundforwomen.org), [GlobalGiving](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.globalgiving.org), [Open Society Foundations](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.opensocietyfoundations.org)) +* **Blocking of specific websites that defend human rights through grants and petitions** ([Change.org](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.change.org), [Global Fund for Women](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.globalfundforwomen.org), [GlobalGiving](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.globalgiving.org), [Open Society Foundations](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2024-01-01&until=2024-04-02&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.opensocietyfoundations.org)) * **Blocking of specific social networking sites** ([Clubhouse](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2023-12-01&until=2024-01-31&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.clubhouse.com) and [4chan](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2023-12-01&until=2024-01-31&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=www.4chan.org)) * **Blocking of a specific VPN** ([ProtonVPN](https://explorer.ooni.org/chart/mat?probe_cc=TZ&since=2023-12-01&until=2024-01-31&time_grain=day&axis_x=measurement_start_day&test_name=web_connectivity&domain=protonvpn.com)) diff --git a/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.sw.md b/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.sw.md index 47f45747a..de0968386 100644 --- a/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.sw.md +++ b/content/post/2024-tanzania-lgbtiq-censorship-and-other-targeted-blocks/index.sw.md @@ -3,7 +3,7 @@ title: "Tanzania: Ongezeko la udhibiti wa LGBTIQ mtandaoni na vizuizi vingine" description: "Ripoti hii imeonyesha udhibiti wa tovuti za LGBTIQ na aina nyingine za udhibiti zilizokusudiwa nchini Tanzania kutokana na uchambuzi wa OONI data." author: "Maria Xynou (OONI), Mehul Gulati (OONI), Tori Fransis" date: "2024-04-11" -tags: ["country-tz"] +tags: ["country-tz", "theme-human_rights", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/belarus-fries-onion.md b/content/post/belarus-fries-onion.md index 2f2e66dad..10e717979 100644 --- a/content/post/belarus-fries-onion.md +++ b/content/post/belarus-fries-onion.md @@ -2,13 +2,13 @@ title: "urandom.pcap: Belarus (finally) bans Tor" author: "Leonid Evdokimov" date: "2016-12-08" -tags: ["belarus", "TCP injections", "country-by"] +tags: ["belarus", "TCP injections", "country-by", "theme-circumvention"] categories: ["report"] --- **Country:** Belarus -**Probed ISPs:** Beltelecom ([AS 6697](https://stat.ripe.net/AS6697)) +**Probed ISPs:** Beltelecom ([AS 6697](https://stat.ripe.net/AS6697)) **Censorship method:** TCP injections diff --git a/content/post/brazil-whatsapp-block.md b/content/post/brazil-whatsapp-block.md index 1452a29b7..6996f4ae0 100644 --- a/content/post/brazil-whatsapp-block.md +++ b/content/post/brazil-whatsapp-block.md @@ -2,7 +2,7 @@ title: "OONI Data Reveals How WhatsApp Was Blocked (Again) in Brazil" author: "Vasilis Ververis, Maria Xynou, Will Scott" date: "2016-05-06" -tags: ["brazil", "whatsapp", "country-br"] +tags: ["brazil", "whatsapp", "country-br", "theme-im"] categories: ["report"] aliases: - /whatsapp-blocked-in-brazil-again diff --git a/content/post/brazil-whatsapp-block.pt.md b/content/post/brazil-whatsapp-block.pt.md index 70c29b38d..7fa399b9e 100644 --- a/content/post/brazil-whatsapp-block.pt.md +++ b/content/post/brazil-whatsapp-block.pt.md @@ -2,7 +2,7 @@ title: "WhatsApp bloqueado (outra vez) no Brasil" author: "Vasilis Ververis, Maria Xynou, Will Scott" date: "2016-05-06" -tags: ["brazil", "whatsapp", "country-br"] +tags: ["brazil", "whatsapp", "country-br", "theme-im"] categories: ["report"] aliases: - /whatsapp-blocked-in-brazil-again-pt diff --git a/content/post/cuba-internet-censorship-2017.md b/content/post/cuba-internet-censorship-2017.md index 1de006d88..572169b4f 100644 --- a/content/post/cuba-internet-censorship-2017.md +++ b/content/post/cuba-internet-censorship-2017.md @@ -2,7 +2,7 @@ title: "Measuring Internet Censorship in Cuba's ParkNets" author: "Maria Xynou, Arturo Filastò, Simone Basso" date: "2017-08-28" -tags: ["cuba", "censorship", "research-report", "country-cu"] +tags: ["cuba", "censorship", "research-report", "country-cu", "theme-news_media", "theme-human_rights", "theme-circumvention", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/cuba-referendum.md b/content/post/cuba-referendum.md index 812fab7ee..eecf48b0f 100644 --- a/content/post/cuba-referendum.md +++ b/content/post/cuba-referendum.md @@ -2,7 +2,7 @@ title: "Cuba blocks independent media amid 2019 constitutional referendum" author: "Eduardo Ernesto Pérez Pujol, Arturo Filastò (OONI), Maria Xynou (OONI)" date: "2019-02-26" -tags: ["cuba", "censorship", "country-cu"] +tags: ["cuba", "censorship", "country-cu", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/egypt-censors.md b/content/post/egypt-censors.md index 395d28f7f..f78056bca 100644 --- a/content/post/egypt-censors.md +++ b/content/post/egypt-censors.md @@ -2,7 +2,7 @@ title: "#EgyptCensors: Evidence of recent censorship events in Egypt" author: "Maria Xynou (OONI), Vasilis Ververis (OONI), Arturo Filastò (OONI), Wafa Ben Hassine (Access Now)" date: "2017-06-19T15:30:00Z" -tags: ["egypt", "censorship", "country-eg"] +tags: ["egypt", "censorship", "country-eg", "theme-news_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/egypt-internet-censorship.md b/content/post/egypt-internet-censorship.md index 78e79878b..21cbfc872 100644 --- a/content/post/egypt-internet-censorship.md +++ b/content/post/egypt-internet-censorship.md @@ -2,7 +2,7 @@ title: "The State of Internet Censorship in Egypt" author: "Leonid Evdokimov (OONI), Maria Xynou (OONI), Mohammad El-Taher (AFTE), Hassan Al-Azhary (AFTE), Sarah Mohsen (AFTE)" date: "2018-07-02" -tags: ["egypt", "censorship", "research-report", "country-eg"] +tags: ["egypt", "censorship", "research-report", "country-eg", "theme-news_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/egypt-network-interference.md b/content/post/egypt-network-interference.md index a28d158c7..5b02600d5 100644 --- a/content/post/egypt-network-interference.md +++ b/content/post/egypt-network-interference.md @@ -2,7 +2,7 @@ title: "Egypt: Media censorship, Tor interference, HTTPS throttling and ads injections?" author: "Leonid Evdokimov, Vasilis Ververis" date: "2016-10-27" -tags: ["egypt", "DPI", "TCP injections", "network throttling", "country-eg"] +tags: ["egypt", "DPI", "TCP injections", "network throttling", "country-eg", "theme-news_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/ethiopia-report.md b/content/post/ethiopia-report.md index b7298e64f..513070502 100644 --- a/content/post/ethiopia-report.md +++ b/content/post/ethiopia-report.md @@ -2,7 +2,7 @@ title: "Ethiopia: Evidence of social media blocking and internet censorship" author: "Maria Xynou (OONI), Arturo Filastò (OONI), Amnesty International, Girma Walabuma" date: "2016-12-14" -tags: ["ethiopia", "DPI", "protests", "whatsapp", "censorship", "research-report", "country-et"] +tags: ["ethiopia", "DPI", "protests", "whatsapp", "censorship", "research-report", "country-et", "theme-social_media", "theme-im", "theme-human_rights", "theme-circumvention", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/ethiopia-unblocking.md b/content/post/ethiopia-unblocking.md index 77f5dc47c..8a527ca0e 100644 --- a/content/post/ethiopia-unblocking.md +++ b/content/post/ethiopia-unblocking.md @@ -2,7 +2,7 @@ title: "Ethiopia: Verifying the unblocking of websites" author: "Berhan Taye (Access Now), Maria Xynou (OONI), Leonid Evdokimov (OONI), Moses Karanja (University of Toronto)" date: "2018-06-29T19:49:00" -tags: ["ethiopia", "unblocking", "country-et"] +tags: ["ethiopia", "unblocking", "country-et", "theme-news_media", "theme-social_media", "theme-circumvention", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/ethiopia-whatsapp-telegram.md b/content/post/ethiopia-whatsapp-telegram.md index 1aa1977f6..1c38388eb 100644 --- a/content/post/ethiopia-whatsapp-telegram.md +++ b/content/post/ethiopia-whatsapp-telegram.md @@ -2,11 +2,11 @@ title: "Ethiopia: From internet blackouts to the blocking of WhatsApp and Telegram" author: "Maria Xynou (OONI), Arturo Filastò (OONI), Moses Karanja (University of Toronto)" date: "2019-06-21" -tags: ["ethiopia", "DPI", "protests", "telegram", "whatsapp", "censorship", "country-et"] +tags: ["ethiopia", "DPI", "protests", "telegram", "whatsapp", "censorship", "country-et", "theme-im", "theme-social_media"] categories: ["report"] --- -Communication access in Ethiopia has significantly been disrupted over the last week. Internet access and Short Message Services (SMS) were [disconnected](https://web.archive.org/save/https://twitter.com/addisstandard/status/1139127606211727360) intermittently across the country.   +Communication access in Ethiopia has significantly been disrupted over the last week. Internet access and Short Message Services (SMS) were [disconnected](https://web.archive.org/save/https://twitter.com/addisstandard/status/1139127606211727360) intermittently across the country. During the first half of the week, internet connectivity was [shut down completely](https://ioda.caida.org/ioda/dashboard#view=inspect&entity=country/ET&lastView=overview&from=1560166834&until=now). After access was restored, both [WhatsApp](https://explorer.ooni.io/measurement/20190615T215418Z_AS24757_NKzIEujAd7rEN95SsIZnraLm1yjqdU79EqczVKFLUzsOJh4DLE) and [Telegram](https://explorer.ooni.io/measurement/20190615T222820Z_AS24757_4flsDD8gZDpmh2jr60nmgLuGGymkNsRA8HOejGDOFZZDEWr6EP) were blocked. A few days ago, Ethio Telecom (the only telecom in the country) [published a notice](https://web.archive.org/web/20190619122350/https:/twitter.com/ethiotelecom_/status/1141013304762146817), acknowledging the internet disruptions, but came short on details about the extent of communication disruptions or why they occurred in the first place. diff --git a/content/post/how-pakistan-blocked-social-media.md b/content/post/how-pakistan-blocked-social-media.md index ac6de7779..73442091f 100644 --- a/content/post/how-pakistan-blocked-social-media.md +++ b/content/post/how-pakistan-blocked-social-media.md @@ -2,7 +2,7 @@ title: "How Pakistan blocked news outlets, social media sites, and IM apps amidst protests" author: "Maria Xynou (OONI), Arturo Filastò (OONI), Shahzad Ahmad (Bytes for All Pakistan), Abdul Salam (Bytes for All Pakistan)" date: "2017-11-29" -tags: ["pakistan", "censorship", "country-pk"] +tags: ["pakistan", "censorship", "country-pk", "theme-news_media", "theme-social_media", "theme-im"] categories: ["report"] --- diff --git a/content/post/indonesia-internet-censorship.md b/content/post/indonesia-internet-censorship.md index a112ca8df..6be356f99 100644 --- a/content/post/indonesia-internet-censorship.md +++ b/content/post/indonesia-internet-censorship.md @@ -2,7 +2,7 @@ title: "The State of Internet Censorship in Indonesia" author: "Kay Yen Wong (Sinar Project), Maria Xynou (OONI), Arturo Filastò (OONI), Khairil Yusof (Sinar Project),Tan Sze Ming (Sinar Project)" date: "2017-05-23" -tags: ["indonesia", "censorship", "research-report", "country-id"] +tags: ["indonesia", "censorship", "research-report", "country-id", "theme-news_media", "theme-human_rights", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/iran-internet-censorship.md b/content/post/iran-internet-censorship.md index 7e57f8026..09931bc3f 100644 --- a/content/post/iran-internet-censorship.md +++ b/content/post/iran-internet-censorship.md @@ -2,7 +2,7 @@ title: "Internet Censorship in Iran: Network Measurement Findings from 2014-2017" author: "Maria Xynou (OONI), Arturo Filastò (OONI), Mahsa Alimardani (ARTICLE 19), Sina Kouhi (ASL19), Kyle Bowen (ASL19), Vmon (ASL19), Amin Sabeti (Small Media)" date: "2017-09-28" -tags: ["iran", "censorship", "research-report", "country-ir"] +tags: ["iran", "censorship", "research-report", "country-ir", "theme-human_rights", "theme-news_media", "theme-social_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/jordan-measuring-facebook-interference.md b/content/post/jordan-measuring-facebook-interference.md index 1be0099cf..64fb1e311 100644 --- a/content/post/jordan-measuring-facebook-interference.md +++ b/content/post/jordan-measuring-facebook-interference.md @@ -2,7 +2,7 @@ title: "Jordan: Measuring Facebook live-streaming interference during protests" author: "Issa Mahasneh (Jordan Open Source Association), Simone Basso (OONI)" date: "2019-06-12" -tags: ["country-jo", "research-report"] +tags: ["country-jo", "research-report", "theme-social_media"] categories: ["report"] --- Last December, [anti-austerity protests](https://www.washingtonpost.com/news/monkey-cage/wp/2018/12/20/youth-revolts-and-political-opposition-in-jordan/?utm_term=.80244368c558) erupted in Jordan against a controversial fiscal reform bill. The protests ensued once a week for several weeks thereafter. Amid the protests, locals reported that they were unable to view live-streaming from Facebook. But they also reported that viewing live-streaming was otherwise possible when protests were not taking place. diff --git a/content/post/malaysia-report.md b/content/post/malaysia-report.md index a5f2b4ac4..00862e3a1 100644 --- a/content/post/malaysia-report.md +++ b/content/post/malaysia-report.md @@ -2,7 +2,7 @@ title: "The State of Internet Censorship in Malaysia" author: "Maria Xynou (OONI), Arturo Filastò (OONI), Khairil Yusof (Sinar Project), Tan Sze Ming (Sinar Project)" date: "2016-12-20" -tags: ["malaysia", "censorship", "research-report", "country-my"] +tags: ["malaysia", "censorship", "research-report", "country-my", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/mali-disruptions-amid-2018-election.md b/content/post/mali-disruptions-amid-2018-election.md index 7a531de73..a50c936e3 100644 --- a/content/post/mali-disruptions-amid-2018-election.md +++ b/content/post/mali-disruptions-amid-2018-election.md @@ -2,7 +2,7 @@ title: "Mali: Social media disruptions amid 2018 presidential election?" author: "Maria Xynou (OONI), Julie Owono (Internet Sans Frontieres)" date: "2018-07-31" -tags: ["mali", "censorship", "whatsapp", "socialmedia", "country-ml"] +tags: ["mali", "censorship", "whatsapp", "socialmedia", "country-ml", "theme-social_media", "theme-im", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/pakistan-internet-censorship.md b/content/post/pakistan-internet-censorship.md index b2a33b7ee..a1666ad83 100644 --- a/content/post/pakistan-internet-censorship.md +++ b/content/post/pakistan-internet-censorship.md @@ -2,7 +2,7 @@ title: "Internet Censorship in Pakistan: Findings from 2014-2017" author: "Haroon Baloch (Bytes for All Pakistan), Maria Xynou (OONI), Arturo Filastò (OONI)" date: "2017-10-18" -tags: ["pakistan", "censorship", "research-report", "country-pk"] +tags: ["pakistan", "censorship", "research-report", "country-pk", "theme-human_rights"] categories: ["report"] --- diff --git a/content/post/parknet-short-documentary-cuba.md b/content/post/parknet-short-documentary-cuba.md index 94a8cb1b7..1dd2e9b1a 100644 --- a/content/post/parknet-short-documentary-cuba.md +++ b/content/post/parknet-short-documentary-cuba.md @@ -2,7 +2,7 @@ title: "ParkNet: Short Documentary on Internet Censorship in Cuba" author: "Arturo Filastò, Maria Xynou, Simone Basso" date: "2018-04-23" -tags: ["cuba", "censorship", "documentary", "country-cu"] +tags: ["cuba", "censorship", "documentary", "country-cu", "theme-news_media"] categories: ["blog"] --- diff --git a/content/post/resurgence-internet-censorship-ethiopia-2019.md b/content/post/resurgence-internet-censorship-ethiopia-2019.md index 0adb1d91c..2145c828e 100644 --- a/content/post/resurgence-internet-censorship-ethiopia-2019.md +++ b/content/post/resurgence-internet-censorship-ethiopia-2019.md @@ -2,7 +2,7 @@ title: "Resurgence of Internet Censorship in Ethiopia: Blocking of WhatsApp, Facebook, and African Arguments" author: "Maria Xynou (OONI), Moses Karanja (University of Toronto), Berhan Taye (Access Now), Arturo Filastò (OONI)" date: "2019-08-14" -tags: ["country-et"] +tags: ["country-et", "theme-social_media", "theme-im", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/south-sudan-censorship.md b/content/post/south-sudan-censorship.md index bb292d1e2..d7f266030 100644 --- a/content/post/south-sudan-censorship.md +++ b/content/post/south-sudan-censorship.md @@ -2,7 +2,7 @@ title: "South Sudan: Measuring Internet Censorship in the World's Youngest Nation" author: "Maria Xynou (OONI), Kenyi Yasin Abdallah Kenyi (TAHURID), Leonid Evdokimov (OONI), Stanley Nyombe Gore (TAHURID)" date: "2018-08-01" -tags: ["south sudan", "censorship", "country-ss"] +tags: ["south sudan", "censorship", "country-ss", "theme-news_media"] categories: ["report"] --- diff --git a/content/post/t-mobile-usa-web-guard.md b/content/post/t-mobile-usa-web-guard.md index 71d9be8f5..839a294b7 100644 --- a/content/post/t-mobile-usa-web-guard.md +++ b/content/post/t-mobile-usa-web-guard.md @@ -44,6 +44,7 @@ error page how to do so. This means that T-Mobile USA customers may be drawn into the Filternet without knowing how to get out. ### How is the censorship performed? + The technique being used by T-Mobile USA to restrict access to certain content involves deploying a transprent HTTP proxy. This proxy intercepts every HTTP packet going through the network and checks it's Host header field to see if diff --git a/content/post/thailand-internet-censorship.md b/content/post/thailand-internet-censorship.md index 4d15dd5da..729d444f5 100644 --- a/content/post/thailand-internet-censorship.md +++ b/content/post/thailand-internet-censorship.md @@ -2,7 +2,7 @@ title: "The State of Internet Censorship in Thailand" author: "Kay Yen Wong (Sinar Project), Maria Xynou (OONI), Arturo Filastò (OONI), Khairil Yusof (Sinar Project),Tan Sze Ming (Sinar Project), Thai Netizen Network" date: "2017-03-20" -tags: ["thailand", "censorship", "research-report", "country-th"] +tags: ["thailand", "censorship", "research-report", "country-th", "theme-news_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/uganda-social-media-blocked.md b/content/post/uganda-social-media-blocked.md index 96a561893..4ffd23668 100644 --- a/content/post/uganda-social-media-blocked.md +++ b/content/post/uganda-social-media-blocked.md @@ -2,7 +2,7 @@ title: "How Uganda blocked social media, again" author: "Maria Xynou, Arturo Filastò" date: "2016-05-17" -tags: ["uganda", "socialmedia", "country-ug"] +tags: ["uganda", "social-media", "country-ug", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/uganda-social-media-tax.md b/content/post/uganda-social-media-tax.md index b7d55be66..190fd4729 100644 --- a/content/post/uganda-social-media-tax.md +++ b/content/post/uganda-social-media-tax.md @@ -2,7 +2,7 @@ title: "Uganda's Social Media Tax through the lens of network measurements" author: "Maria Xynou (OONI), Leonid Evdokimov (OONI), Arturo Filastò (OONI), DefendDefenders, with artwork and contributions from POLLICY" date: "2018-11-12" -tags: ["uganda", "research-report", "country-ug"] +tags: ["uganda", "research-report", "country-ug", "theme-social_media", "theme-im", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/venezuela-blocking-wikipedia-and-social-media-2019.md b/content/post/venezuela-blocking-wikipedia-and-social-media-2019.md index 4c02af4b5..4ec6b1082 100644 --- a/content/post/venezuela-blocking-wikipedia-and-social-media-2019.md +++ b/content/post/venezuela-blocking-wikipedia-and-social-media-2019.md @@ -2,7 +2,7 @@ title: "From the blocking of Wikipedia to Social Media: Venezuela's Political Crisis" author: "Andrés Azpúrua (Venezuela Inteligente / VEsinFiltro), Mariengracia Chirinos (IPYS Venezuela), Arturo Filastò (OONI), Maria Xynou (OONI), Simone Basso (OONI), Kanishk Karan (Digital Forensic Research Lab)" date: "2019-01-29" -tags: ["venezuela", "censorship", "country-ve"] +tags: ["venezuela", "censorship", "country-ve", "theme-social_media"] categories: ["report"] --- diff --git a/content/post/venezuela-internet-censorship.md b/content/post/venezuela-internet-censorship.md index 8f79bffb3..6a084315a 100644 --- a/content/post/venezuela-internet-censorship.md +++ b/content/post/venezuela-internet-censorship.md @@ -2,7 +2,7 @@ title: "The State of Internet Censorship in Venezuela" author: "Mariengracia Chirinos (IPYS Venezuela), Andrés Azpúrua (Venezuela Inteligente / VEsinFiltro), Leonid Evdokimov (OONI), Maria Xynou (OONI)" date: "2018-08-16" -tags: ["country-ve", "research-report", "censorship"] +tags: ["country-ve", "research-report", "censorship", "theme-news_media", "theme-circumvention"] categories: ["report"] --- diff --git a/content/post/zimbabwe-protests-social-media-blocking-2019.md b/content/post/zimbabwe-protests-social-media-blocking-2019.md index fe5bb5756..986fb42d5 100644 --- a/content/post/zimbabwe-protests-social-media-blocking-2019.md +++ b/content/post/zimbabwe-protests-social-media-blocking-2019.md @@ -2,7 +2,7 @@ title: "Zimbabwe protests: Social media blocking and internet blackouts" author: "Maria Xynou (OONI), Arturo Filastò (OONI), Tawanda Mugari (Digital Society of Zimbabwe), Natasha Msonza (Digital Society of Zimbabwe)" date: "2019-01-23" -tags: ["zimbabwe", "censorship", "country-zw"] +tags: ["zimbabwe", "censorship", "country-zw", "theme-social_media", "theme-im"] categories: ["report"] ---