Skip to content
Theme
OnDevice Tools

SoftwareApps & Productivity

Meta Muse AI agent faces privacy concerns over crossed chats

Meta employees say the Muse AI agent has shown them other people's messages and financial details, according to a Reuters report cited by TechCrunch. Meta had not immediately responded to a request for comment.

Meta Muse AI agent app illustration showing mixed-up private messages

Key takeaways

  • Meta employees reported that Muse showed messages and bank balances from other accounts.
  • TechCrunch cited a Reuters report based on internal company posts.
  • Muse launched in September and has more than 10 million downloads.
  • Meta had not immediately replied to TechCrunch.
In this article
  1. What happened
  2. Reported examples
  3. A fast-growing app
  4. What remains unclear
  5. What readers should know

What happened

The Meta Muse AI agent is facing privacy concerns after employees said it mixed conversations from different accounts and revealed personal information. TechCrunch reported the issue on 19 November 2026, citing a Reuters report that reviewed internal company posts.

Muse is described as an AI agent, a type of software tool that can carry out tasks across apps and services. The employees said the tool appeared to show information that belonged to other people. The available material does not provide a technical explanation.

TechCrunch said Meta had not replied immediately when asked for comment. The report therefore relies on employee accounts and internal posts, rather than a statement from the company. That lack of comment leaves the company's position unknown.

Reported examples

Illustration: messaging, banking, AI, Meta logo and the labels AI Agent, Meta, Muse

The employees gave several examples of the reported problem. They said Muse showed messages about childcare requests, career advice, product orders and bank balances that they did not recognise. The examples were attributed to employees, not to Meta.

In one case, an employee said a chat about weather and tide information included an unexpected message. The message reportedly included two bank updates, according to TechCrunch. It listed balances for Capital One and Charles Schwab.

In another case, an employee said Muse reminded her, on her personal phone, to answer a message from her sister on Signal. Signal is a messaging app known for encryption. The employee said she had not given permission for access to other apps.

A fast-growing app

The report says Muse launched in September. It also says the app has more than 10 million downloads. The launch month and download figure come from TechCrunch.

TechCrunch said the app has grown very quickly. That rapid growth makes the reported privacy concerns more significant. The download figure shows the app has reached a wide audience.

The available material does not say whether the problem affects all users. It also does not say whether the reported cases involve a small number of accounts or a broader fault. Because the examples involve messages and bank balances, the issue goes beyond a simple user interface error.

The reported examples include financial details, which are often sensitive. They also include messages that may be personal or confidential.

What remains unclear

The Reuters report looked at internal company posts. Those posts are the basis for the employee accounts, rather than a confirmed technical finding from Meta. The report's account is based on what employees posted internally.

The material does not identify the cause of the mix-ups. It does not say whether Meta has fixed the issue, notified users or published an official advisory. Without those details, the scale of the issue remains uncertain.

It also does not say whether the reported data was shown to the wrong person, stored incorrectly or retrieved by mistake. Those details would be needed to understand the scope of the problem. The report does not provide a timeline.

What readers should know

No mitigation steps or official advisory were included in the available material. Readers should look for official guidance from Meta for confirmed details and any recommended steps. No fix, workaround or safety notice was described in the material.

The employee accounts remain claims from internal posts. They have not been confirmed by Meta, and the report does not provide evidence that every user is affected. The available material does not say whether the issue has been resolved.

The report does not say whether users can disable access between apps. It also does not say whether Meta has contacted the people whose data appeared.

For now, the story is a privacy warning about an AI agent with a large user base. It also raises questions about permission and oversight for access between apps.

Sources

  1. Meta’s AI agent Muse is mixing up conversations and exposing private data, report saysTechCrunch

Tags

Spotted a mistake? See our corrections policy. Read how News by OnDevice works in our editorial policy.