diff --git a/litellm/proxy/guardrails/guardrail_hooks/bedrock_guardrails.py b/litellm/proxy/guardrails/guardrail_hooks/bedrock_guardrails.py
index 62c997659b..fb14ccce50 100644
--- a/litellm/proxy/guardrails/guardrail_hooks/bedrock_guardrails.py
+++ b/litellm/proxy/guardrails/guardrail_hooks/bedrock_guardrails.py
@@ -605,13 +605,13 @@ class BedrockGuardrail(CustomGuardrail, BaseAWSLLM):
         """
         Only raise exception for "BLOCKED" actions, not for "ANONYMIZED" actions.
 
-        If `self.mask_request_content` or `self.mask_response_content` is set to `True`, 
-        then use the output from the guardrail to mask the request or response content.
-        
-        However, even with masking enabled, content with action="BLOCKED" should still 
-        raise an exception, only content with action="ANONYMIZED" should be masked.
+        If `self.mask_request_content` or `self.mask_response_content` is set to `True`, then use the output from the guardrail to mask the request or response content.
         """
 
+        # if user opted into masking, return False. since we'll use the masked output from the guardrail
+        if self.mask_request_content or self.mask_response_content:
+            return False
+
         # if no intervention, return False
         if response.get("action") != "GUARDRAIL_INTERVENED":
             return False
diff --git a/tests/test_litellm/proxy/guardrails/guardrail_hooks/test_bedrock_guardrails.py b/tests/test_litellm/proxy/guardrails/guardrail_hooks/test_bedrock_guardrails.py
index 84d320a0a2..69b0bb27b4 100644
--- a/tests/test_litellm/proxy/guardrails/guardrail_hooks/test_bedrock_guardrails.py
+++ b/tests/test_litellm/proxy/guardrails/guardrail_hooks/test_bedrock_guardrails.py
@@ -1101,91 +1101,3 @@ async def test_bedrock_apply_guardrail_with_only_tool_calls_response():
         # Verify that the Bedrock API was NOT called since there's no text to process
         mock_api_request.assert_not_called()
         print("✅ apply_guardrail with tool_calls test passed - no API call made")
-
-
-@pytest.mark.asyncio
-async def test_bedrock_guardrail_blocked_content_with_masking_enabled():
-    """Test that BLOCKED content raises exception even when masking is enabled
-    
-    This test verifies the bug fix where previously mask_request_content=True or
-    mask_response_content=True would bypass all BLOCKED content checks. Now it
-    properly distinguishes between BLOCKED (raise exception) and ANONYMIZED (apply masking).
-    """
-    
-    # Create guardrail with masking enabled
-    guardrail = BedrockGuardrail(
-        guardrailIdentifier="test-guardrail",
-        guardrailVersion="DRAFT",
-        mask_request_content=True,  # Masking enabled
-        mask_response_content=True,  # Masking enabled
-    )
-    
-    # Mock Bedrock response with BLOCKED content (hate speech)
-    blocked_response = {
-        "action": "GUARDRAIL_INTERVENED",
-        "assessments": [
-            {
-                "contentPolicy": {
-                    "filters": [
-                        {
-                            "type": "HATE",
-                            "confidence": "HIGH",
-                            "action": "BLOCKED",  # Should raise exception
-                        }
-                    ]
-                },
-                "sensitiveInformationPolicy": {
-                    "piiEntities": [
-                        {
-                            "type": "NAME",
-                            "match": "John Doe",
-                            "action": "ANONYMIZED",  # Should be masked
-                        }
-                    ]
-                },
-            }
-        ],
-        "outputs": [{"text": "Content blocked due to policy violation"}],
-    }
-    
-    mock_bedrock_response = MagicMock()
-    mock_bedrock_response.status_code = 200
-    mock_bedrock_response.json.return_value = blocked_response
-    
-    # Mock credentials
-    mock_credentials = MagicMock()
-    mock_credentials.access_key = "test-access-key"
-    mock_credentials.secret_key = "test-secret-key"
-    mock_credentials.token = None
-    
-    request_data = {
-        "model": "gpt-4o",
-        "messages": [
-            {"role": "user", "content": "Test message with PII and hate speech"},
-        ],
-    }
-    
-    # Mock AWS-related methods
-    with patch.object(
-        guardrail.async_handler, "post", new_callable=AsyncMock
-    ) as mock_post, patch.object(
-        guardrail, "_load_credentials", return_value=(mock_credentials, "us-east-1")
-    ), patch.object(
-        guardrail, "_prepare_request", return_value=MagicMock()
-    ):
-        mock_post.return_value = mock_bedrock_response
-        
-        # Should raise HTTPException for BLOCKED content
-        with pytest.raises(HTTPException) as exc_info:
-            await guardrail.make_bedrock_api_request(
-                source="INPUT",
-                messages=request_data.get("messages"),
-                request_data=request_data,
-            )
-        
-        # Verify exception details
-        assert exc_info.value.status_code == 400
-        assert "Violated guardrail policy" in str(exc_info.value.detail)
-        
-        print("✅ BLOCKED content with masking enabled raises exception correctly")
-