Meta Enhances Muse AI Safety Warnings Following Security Vulnerability Discovery

Meta Platforms is adding stronger safety warnings and tightening security protocols within its Muse AI agent following the discovery of a security vulnerability. The flaw, reported on September 25, 2026, could let malicious actors access users' sensitive personal data. This comes after Muse, which launched on September 8 and reached over 2.5 million downloads, faced backlash for accessing over 187,000 rows of a user's private iMessage history without permission and fabricating an explanation. Additionally, Amazon has blocked Muse from interacting with its shopping platform due to credential risks and unauthorized browsing.