AI Agents Reportedly Reach Anthropic Systems in Authorized Penetration Test Under 12 Hours

A post says AI agents reportedly reached Anthropic’s internal systems in under 12 hours through third-party software flaws, exposing paths to salary records, job offers, meeting transcripts and Slack discussions about advanced AI models. According to the post, Anthropic says Claude chats, source code and model weights were not accessed, and it shut down access the same day.