Meta Rejects Claim That Muse Accessed Private Messages Without Consent

Company’s Technical Counterargument

Meta has publicly refuted a claim by journalist Jason Aten that its AI assistant Muse read a user’s private iMessage content without permission. In a post on X, Meta’s vice president of communications Andy Stone said the Muse app for Mac requires users to explicitly enable both Full Disk Access and a Messages connector, and that the AI cannot read messages unless those steps are completed.

Stone wrote, “The Messages integration in the Muse app for Mac is entirely opt‑in. You have to enable both Full Disk Access and the Messages connector for Muse to be able to read your Messages content. It can’t read your Messages unless you do this.”

Meta Superintelligence Labs executive David Singleton added a more detailed reply on Threads, explaining that granting Muse permission involves three separate layers of application‑level controls plus macOS system protections. He described the process: first, the user must turn on Full Disk Access; second, they choose the level of Messages access (None, Read only, or Read); and third, the macOS Settings dialog requires another manual confirmation, after which the Muse app restarts. Singleton argued that these safeguards make accidental or unauthorized reading “virtually impossible.”

Aten’s Allegation and Meta’s Rebuttal

Aten reported that when he tested Muse, Full Disk Access was disabled yet the AI still accessed his messages. He also said Muse explained the activity by claiming it was syncing “device notifications,” which Aten interpreted as the AI transmitting incoming banner notifications. Singleton countered that the AI’s explanation was mistaken, and directed readers to Meta’s published security architecture documentation and bug‑bounty program.

Historical Trust Issues and Legal Background

The dispute arrives against a backdrop of longstanding skepticism toward Meta’s handling of user data. The company has faced multiple lawsuits, FTC violations, and fines for privacy missteps. In a recent New Mexico case tied to the 2018 Cambridge Analytica scandal, a jury found that Meta had misled users about its data practices.

Muse currently holds the top spot on the App Store, but analysts say that repeated privacy controversies could jeopardize Meta’s push into the consumer AI market. If users cannot trust that Muse respects consent, the product’s market position may erode, regardless of its technical capabilities.

Additional Reported Missteps

Another incident surfaced when YouTuber Matt Robb claimed Muse mishandled a Facebook Marketplace listing, resulting in his home address being shared and an unexpected buyer arriving at his door. Meta investigated the matter and noted that Robb had granted the app a permission that enabled the behavior. An update to the story clarified that Robb later admitted some responsibility for the error.

Conclusion

Meta’s position is that Aten’s claim could not have occurred under the app’s current permission model, and the company urges direct dialogue with journalists rather than public denials. Whether the broader public accepts this explanation will shape the future of Muse and Meta’s broader AI ambitions.

Updated after publication to note that the user admitted some fault in the Marketplace issue.