False/incorrect information posting on copilot

Copilot responses are generated by AI and are not always accurate or factual. Content about legal or criminal activity that appears associated with a name can therefore be incorrect, and all guidance emphasizes reviewing and not relying on such output as fact.

To address harmful or inaccurate content involving a name:

1. When the problematic response appears, use the built‑in feedback/reporting controls:
   • Select the thumbs‑down or equivalent feedback option on the specific response.
   • In the feedback form, clearly state that the content is false, harmful, and associates a real person’s name with criminal or legal activity.
2. If using Copilot in Microsoft 365 apps or services, repeat this feedback each time similar content appears so it can be reviewed and filtered more effectively.
3. For general Copilot on the web or mobile, use the “Give feedback” or “Report” option (for example, via the flag icon or long‑press menu) on the specific response and describe the issue as clearly as possible.

Copilot includes safeguards and ongoing review processes to reduce harmful or inappropriate content, and user reports are a primary mechanism to have specific problematic outputs reviewed and mitigated.

If the concern is also about privacy and how data is used, review and adjust Copilot privacy controls and history settings for the personal Microsoft account, including deleting conversation history if desired.

References:

• Privacy FAQ for Microsoft Copilot
• Microsoft Copilot privacy controls
• FAQ for Copilot data security and privacy for Dynamics 365 and Power Platform
• Frequently Asked Questions about Copilot in Whiteboard
• Frequently asked questions about Copilot in Loop
• Frequently asked questions about Copilot in the Clipchamp video player