Meta's Muse AI Agent Reported to Mix Up Conversations and Expose Private User Data
Internal reports from Meta employees suggest that the company's rapidly growing AI agent, Muse, is experiencing significant privacy failures. According to a Reuters report, the AI has been mixing up conversations across different user accounts, leading to the exposure of sensitive information such as bank account balances from Capital One and Charles Schwab, as well as private messages from third-party apps like Signal. Despite being one of the fastest-growing applications with over 10 million downloads since its September launch, Muse's technical glitches have raised serious internal alarms regarding data security. Employees reported receiving unsolicited information ranging from incorrect product orders to career advice and personal reminders that they did not authorize or request.