diff --git a/pytorch_lightning/trainer/trainer.py b/pytorch_lightning/trainer/trainer.py
index 62d7deb0eb378..35709c8b73d35 100644
--- a/pytorch_lightning/trainer/trainer.py
+++ b/pytorch_lightning/trainer/trainer.py
@@ -133,7 +133,7 @@ def __init__(
         distributed_backend: Optional[str] = None,
         automatic_optimization: Optional[bool] = None,
         move_metrics_to_cpu: bool = False,
-        enable_pl_optimizer: bool = True,
+        enable_pl_optimizer: bool = False,
     ):
         r"""
         Customize every aspect of training via flags
diff --git a/tests/callbacks/test_callbacks.py b/tests/callbacks/test_callbacks.py
index c00c712bb3b13..070bb4e9f6989 100644
--- a/tests/callbacks/test_callbacks.py
+++ b/tests/callbacks/test_callbacks.py
@@ -33,6 +33,8 @@ def test_trainer_callback_system(torch_save):
         limit_train_batches=3,
         limit_test_batches=2,
         progress_bar_refresh_rate=0,
+        # todo: enabled since internally we wrap the model for optimizer step, this should be fixed
+        enable_pl_optimizer=True
     )
 
     # no call yet