Meta’s AI agent Muse is mixing up conversations and could be exposing private data, report says
CybersecurityInformation SystemsLaw
THE AI ANGLE
Exposing private user data across separate accounts and external appsMeta employees reported that the AI agent Muse mixed up conversations across different user accounts and exposed private data like bank balances. The incidents show cybersecurity, systems, and legal teams how cross-account data leaks and unauthorized app monitoring can occur in agent software.
THE TEACHING ANGLE
Discuss whether pre-release testing justifies an AI agent exposing financial records across accounts and accessing external apps without clear user consent.Read the original at techcrunch.com Generate teaching or study materials
More in Cybersecurity
- Google Donates Agent Substrate and gVisor to the CNCF to Secure AI Agents at ScaleInfoQ · November 20, 2026
- Signal is blocking Gemini and Siri AI from reading your screenThe Verge · November 20, 2026
- Healthcare firms fail to monitor AI tools as shadow AI leads to breachesHealthcare Dive · November 20, 2026
- Major US banks say all AI agent actions are user authorized, even mistakesTechRadar · November 20, 2026
- Study Finds ChatGPT Tracking Tag In 298 Federal Court Cases And 133 SEC FilingsAbove the Law · November 20, 2026