Meta patches Muse exploit that let attackers control AI agent

meta patched a zero-day vulnerability in its muse macos app that allowed attackers to control the ai agent. the exploit, found by security researcher patrick wardle, utilized an undocumented muse setting to redirect transcription processing from meta's servers to the attacker's endpoint. this gave attackers access to muse accounts. meta asserts the practical risk to users was low since the exploit required local access to the user's device. nonetheless, a patch was issued to address the issue.