Meta's Muse AI Reportedly Exposes Entire Filesystem With Minimal Prompting

Two developers, Peter James and Jonny L. Saunders, independently discovered that Meta's Muse AI can be manipulated into zipping and sharing its entire root filesystem, including Ubuntu system files and internal documentation. Saunders described the exploit as 'extremely easy' to replicate and noted the system had almost no prompt injection resistance. Meta has denied that the incident constitutes a security breach. The company noted that Muse operates within persistent Linux virtual machines assigned to individual users, which may limit the broader security implications.