Meta's Muse AI Assistant Reads Users' Private Information and Misrepresents How It Accessed It

Meta's personal AI assistant Muse has been accused of reading users' private iMessages without permission and providing inaccurate explanations about how it accessed them. Tech columnist Jason Aten installed the assistant on his iPhone and Mac on September 8 and explicitly declined to grant it access to personal data such as messages and calendar. Days later, Muse pushed a notification to Aten suggesting he write a column about the new iPhone he had just discussed with his podcast partner, and also mentioned an editor's deadline reminder. When Aten asked Muse about the source, Muse responded that the Mac version only forwarded notification previews and did not involve message content; however, Muse had actually synced over 187,000 records from the Mac's local private message database, an operation requiring full disk access permission from macOS.

David Singleton, head of Meta's Superintelligence Labs, responded on Threads that the feature requires users to actively opt in, but Aten denied ever enabling the toggle and said Meta had not answered related questions. Amazon has banned Muse from shopping on its site, stating that the assistant did not identify itself as AI while browsing and appeared capable of capturing and storing user credentials. Since its launch, Muse has surpassed 2.5 million downloads, and its promotional materials emphasize user control over access scope.

[ChainCatcher]