Meta has pushed back against a report by Jason Aten, an Inc. columnist, that claimed the company’s AI assistant Muse read a journalist’s private Messages without consent. Andy Stone, Meta’s Vice President of Communications, stated on X that the company does not believe its product accessed Messages without user permission.
Stone emphasized that “the Messages integration in the Muse app for Mac is entirely opt‑in,” and that a user must enable both Full Disk Access and the Messages connector for Muse to be able to read Messages content. “It can’t read your Messages unless you do this,” he wrote.
Company explanation: multi‑step permissions and macOS safeguards
A more technical response came from David Singleton, an executive at Meta Superintelligence Labs, who replied directly to Jason Aten on Threads. Singleton explained that the set of permissions required for Muse to read Messages on Mac involves “three separate steps of application‑level permissions and built‑in macOS system‑level protections,” which he said “can’t be circumvented even if the Muse application had a bug.”
According to Singleton, the flow requires explicitly granting Muse Full Disk Access, after which the user can select what level of access Muse has to the Messages app (None, Read only, or Read). If Full Disk Access is not enabled, those options are grayed out. He added that enabling Full Disk Access invokes the macOS Settings UI and requires the user to manually confirm the action; the process triggers a full restart of the Muse app, making an accidental, unnoticed enabling of the permission unlikely.
The allegation and Meta’s rebuttal
Aten’s report claimed that when Muse read his messages, Full Disk Access was turned off. He also said that when he asked Muse how this happened, the AI replied that it was syncing his “device notifications,” which led Aten to conclude that Muse had been relaying the text of incoming banner notifications on his Mac to the AI agent.
Singleton disputed that account as well, saying the AI was confused and gave an incorrect explanation. He pointed readers to Meta’s information about Muse’s security architecture and its bug bounty process.
In short, Meta’s position is that the scenario Aten described did not and could not have occurred as reported.
Trust and past controversies
Despite Meta’s denial, many remain skeptical of the company’s account — a reaction shaped by Meta’s history of data‑handling controversies. The company has faced lawsuits, FTC findings, and fines related to prior incidents. The report recalls that, just days ago, a New Mexico jury found Meta had misled users about its data practices in a case tied to the 2018 Cambridge Analytica data breach.
Whether users will trust Muse is likely to be a determining factor in Meta’s success in the consumer AI market. Muse is currently performing well and remains No. 1 on the App Store, but Meta’s reputation could suffer if more reports like this surface, true or not. The article’s author suggested Meta should engage directly with the journalist to investigate how the event could have happened, rather than only issuing denials.
Other reported Muse incidents
This is not the only reported Muse overstep. YouTuber Matt Robb recently said Muse mishandled a Facebook Marketplace transaction, resulting in his address being shared and a buyer showing up when Robb was not home. Singleton has indicated via Threads that the company is investigating that report, suggesting Meta thinks that incident might be its fault.
