Meta’s Muse AI Agent Has Been Downloaded Half a Million Times — With a Serious Bug
The security flaw has been patched, but there are plenty of reasons to hesitate about giving AI too much access.
Meta's Muse AI Agent, an AI bot designed as a digital assistant, has downloaded over half a million times in its first week according to internal data. However, Meta faced criticism after a cybersecurity expert discovered a serious zero-day bug in Muse's code. Patrick Wardle, the security researcher, demonstrated that a malicious actor could trick Muse into sending a user's dictations and access tokens to their own server, not Meta's servers.
The head of Meta's Superintelligence Labs, David Singleton, confirmed the patch and emphasized the importance of transparency in privacy and security. While Muse has seen success in the crowded AI agent market, concerns about privacy and security remain high. Tech users are left to question whether Meta can build agentic AI responsibly, especially with recent security issues and dwindling public trust in AI companies.
Written by urgent.news from CNET's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
