Meta employees have raised concerns that its AI agent Muse is mixing up conversations across accounts and surfacing other people’s private data, according to a Reuters report based on internal Meta posts, as relayed by TechCrunch on 19 November 2026. The examples include a chat in which an employee’s Muse surfaced balances from bank accounts that were not theirs, along with childcare requests, product orders and career advice that other people had made. TechCrunch says the reporting describes employees flagging other users’ data appearing in their own chats; it does not establish that employees were deliberately viewing other users’ content.

What Meta said

Meta told TechCrunch by email: “Employee testing is core to the product development process.” It said employees have access to different versions of Muse than the public, TechCrunch reports.

Background

Muse is Meta’s AI agent, which launched in September, according to TechCrunch. Lens has not verified the internal posts Reuters describes.

Sources and notes

Lens read TechCrunch, 19 November 2026, by Aisha Malik, through an automated page summary. The Reuters report itself could not be opened, so its details are as relayed by TechCrunch. Lens did not open Meta’s statement.

Image: a smartphone illustration, used as a stand-in. It is not Meta’s app. Illustration by Santeri Viinamäki, CC BY-SA 4.0, via Wikimedia Commons.