Meta bolsters Muse safety warning after security vulnerability found, The Information reports

Meta Platforms is reinforcing a safety warning for Muse, its personal AI agent, following the discovery of a security vulnerability by a researcher, according to The Information. The flaw, reported via Meta's bug bounty program, could enable an attacker to access a user's sensitive personal information stored in a dedicated virtual machine, which includes emails, files, and other data, as per an internal Meta incident report.

Initially rated as a SEV-2, the vulnerability is classified as Meta's third-highest severity level, typically used for incidents with significant impact. Muse, launched recently, is an AI agent designed for tasks like shopping, travel booking, emailing, and payments. It has quickly gained popularity, topping Apple's App Store and Google Play rankings in free-app charts, with an estimated 2.8 million downloads in its first two weeks. Meta has not yet commented on the issue.