Trendora
Feed
9Ars Technica AI·security

Microsoft Copilot leak exposes hidden prompt bypass

AI summary

Researchers at Varonis used Microsoft 365 Copilot itself to uncover an undocumented prompt parameter that bypassed user-confirmation safeguards. They then demonstrated a path to exfiltrate sensitive data, highlighting a serious security weakness in the enterprise assistant's guardrail design.

In-depth analysis

AI-generated, audience-specific — grounded in this story.

Technologies in this story

Discussion

No comments yet. Start the discussion.