Apple Updates macOS Security to Restrict AI Assistant Access
Apple is tightening its privacy controls following concerns over AI access to private data.
Apple has announced changes to its macOS privacy settings. The company’s decision aims to prevent third-party developers and artificial intelligence (AI) assistants from exploiting system permissions to access user Apple Messages history. This move follows a controversy surrounding Meta’s new AI assistant, Muse.
The Meta AI Assistant Controversy
The issue came to light after technology commentator Jason Aten reported that Meta’s general-purpose AI assistant, Muse, sent him a message containing details from his private Apple Messages correspondence. Aten stated that he had never granted Muse permission to read his private messages and assumed such data remained inaccessible to the AI.
This incident sparked widespread discussion on social media. Users have compared granting AI assistants access to calendars, emails, messages, and online accounts to using a chainsaw—the tools are useful, but they can cause serious harm if handled without caution.
System Loopholes and Apple’s Response
Meta’s Chief Technology Officer (CTO) David Singleton addressed the allegations, explaining that for Muse to access messages, a user would have to manually enable two settings: the macOS system permission for Full Disk Access and the Messages connector within the Muse application itself.
Although users formally granted disk access, most were unaware that Full Disk Access automatically included the Apple Messages database. To address this confusion and prevent potential misuse, Apple has decided to modify how Full Disk Access functions, ensuring that AI assistants can no longer silently read private correspondence.