Meta Muse AI Chat Sharing Leaks Thousands of Private Conversations

Meta's Muse, the company's new general-purpose AI agent, is exposing sensitive user information through its conversation-sharing feature, researchers from security firm Claroc reported Monday. The flaw stems from a design decision in how Muse constructs the URLs used for sharing, making the links easy to guess and discover. This contradicts Meta CEO Mark Zuckerberg's claim last month that Muse was "built from the ground up for privacy and security."

The researchers were able to enumerate 6,800 links that Muse users created for sharing conversations, and after opening a few hundred, they found clearly private information. Exposed material included a candidate's notes for a confidential executive-search call with a frontier AI lab, a marketing company's internal revenue review marked for its CEO, and a small-business owner's loan application with the requested amount.

Other leaked content included a to-do list with a bank fraud alert and partial account number, a real-estate investor's internal performance report with exact financial figures, and a workplace's full-year staff vacation calendar. The researchers noted the effort required was minimal. Meta says it is suspending the sharing feature while it reviews the behavior and considers changes.

Read More at the original source →