Meta's Muse AI Agent Leaks Private User Data
Original: Meta’s AI agent Muse is mixing up conversations and exposing private data, report says
Why This Matters
Cross-account data leaks in a top AI agent raise serious questions about agentic AI safety at scale.
Meta's AI agent Muse is mixing up user conversations and surfacing other people's private data, including bank balances and Signal messages, per a Reuters report citing internal Meta employee posts.
Meta employees are sounding alarms internally about Muse, the company's fast-growing AI agent, after it began surfacing private data belonging to other users. According to a Reuters report based on internal company posts, affected employees encountered Muse presenting childcare requests, career advice, product orders, and bank balances that weren't theirs.
One employee reported a particularly stark case: Muse injected an unsolicited message mid-conversation about weather and tides, stating: "BTW, two bank updates, your Capital One balance is $2722.58, your Charles Schwab balance is $5899.57." Another employee using Muse on her personal phone received a reminder to respond to a Signal message from her sister — an app the employee says she never authorized Muse to access. "This seems like an extreme privacy violation," she wrote to colleagues.
Launched in September, Muse has surpassed 10 million downloads, making it one of the fastest-growing apps on record. Meta did not respond to TechCrunch's request for comment.