Responsible AI Error

Hello Matthew Wagner-Yau,

Thank you for sharing these details. Since even a basic prompt such as “hello” is being blocked, the issue may not be caused by the user’s message itself. The Responsible AI system can also evaluate the agent’s instructions, knowledge content, actions, connector outputs, and the response being generated.

If this agent was created in Microsoft Copilot Studio, please check the test pane or activity map for an error code such as ContentFiltered. You can also try the following:

• Test the agent after temporarily removing its knowledge sources and actions.
• Review the agent’s name, description, and instructions for wording that could be misinterpreted by the safety filter.
• Check the conversation transcript to identify the content processed immediately before the error.
• If content moderation settings are available, review whether the current filtering level is appropriate for the intended business scenario.

Could you share the platform used to build the agent and the exact error code shown in the test pane? Please remove any personal, customer, or organization-specific information before posting screenshots or logs. Thank you so much for your understanding.     

Note: Please follow the steps in our documentation to enable e-mail notifications if you want to receive the related email notification for this thread.