Meta's Muse AI assistant accesses Mac messages despite no permission, raising privacy concerns

Meta’s new AI assistant, Muse, has drawn attention after a user reported that the Mac‑app version appears to retrieve information from personal data sources without explicit permission. Contributing editor Jason Aten of Inc Magazine posted screenshots on Threads showing Muse asking him details about a conversation taking place in the Messages app. Aten noted that he had never granted the assistant access to his messages, yet Muse responded that it had “seen the notification previews,” implying the software was pulling content from the operating system’s notification center rather than from a direct integration.

The incident highlights ongoing concerns about privacy and data handling in consumer AI tools. Muse, which can integrate with Messages, Calendar and Notes on macOS, is marketed as a productivity assistant, but the lack of transparency around how it accesses and processes user information has prompted scrutiny from both users and privacy advocates. Meta has not yet issued a detailed comment, and the story was first reported by The Verge.