Meta Muse Privacy and Security

Muse is Meta’s personal AI agent, designed to work with connected applications and personal information to perform tasks such as completing forms, handling email, and aggregating content.

Its privacy and security questions include the breadth of requested access, permission and revocation controls, and whether user interactions are used for AI training unless users opt out. Meta describes safeguards such as an isolated environment and approval controls for internet interactions, while reports have described personal information being disclosed and message content accessed through notification previews despite denied direct permission. These issues make consent, transparency, and reliable access controls central to understanding Muse.