diff --git a/src/Microsoft.ML.Auto/PipelineSuggesters/PipelineSuggester.cs b/src/Microsoft.ML.Auto/PipelineSuggesters/PipelineSuggester.cs
index 9aa55ceebd4..e9a58f8e74d 100644
--- a/src/Microsoft.ML.Auto/PipelineSuggesters/PipelineSuggester.cs
+++ b/src/Microsoft.ML.Auto/PipelineSuggesters/PipelineSuggester.cs
@@ -221,7 +221,8 @@ private static IEnumerable<SuggestedTransform> CalculateTransforms(
             // this is a work-around for ML.NET bug tracked by https://github.com/dotnet/machinelearning/issues/1969
             if (task == TaskKind.MulticlassClassification)
             {
-                var transform = ValueToKeyMappingExtension.CreateSuggestedTransform(context, DefaultColumnNames.Label, DefaultColumnNames.Label);
+                var labelColumn = columns.First(c => c.Item3 == ColumnPurpose.Label).Item1;
+                var transform = ValueToKeyMappingExtension.CreateSuggestedTransform(context, labelColumn, labelColumn);
                 transforms.Add(transform);
             }
             return transforms;
diff --git a/src/Microsoft.ML.Auto/TrainerExtensions/MultiTrainerExtensions.cs b/src/Microsoft.ML.Auto/TrainerExtensions/MultiTrainerExtensions.cs
index d7ef50f2d7c..2717c87636a 100644
--- a/src/Microsoft.ML.Auto/TrainerExtensions/MultiTrainerExtensions.cs
+++ b/src/Microsoft.ML.Auto/TrainerExtensions/MultiTrainerExtensions.cs
@@ -24,7 +24,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)
@@ -46,7 +46,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)
@@ -89,7 +89,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)
@@ -132,7 +132,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)
@@ -154,7 +154,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)
@@ -176,7 +176,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)
@@ -198,7 +198,7 @@ public ITrainerEstimator CreateInstance(MLContext mlContext, IEnumerable<Sweepab
             ColumnInformation columnInfo)
         {
             var binaryTrainer = _binaryLearnerCatalogItem.CreateInstance(mlContext, sweepParams, columnInfo) as ITrainerEstimatorProducingFloat;
-            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer);
+            return mlContext.MulticlassClassification.Trainers.OneVersusAll(binaryTrainer, labelColumn: columnInfo.LabelColumn);
         }
 
         public PipelineNode CreatePipelineNode(IEnumerable<SweepableParam> sweepParams, ColumnInformation columnInfo)