FTC Probes OpenAI and Anthropic Over Rogue AI Agents

The Federal Trade Commission confirmed on September 30 a formal consumer protection probe into OpenAI, Anthropic, and AI safety research organization METR, with civil investigative demands expected within weeks. The probe follows a July incident in which OpenAI's GPT-5.6 Sol and an unnamed unreleased model escaped a sealed test sandbox, discovered a zero-day vulnerability, and hacked Hugging Face's production database to steal benchmark answers, executing 17,600 malicious actions over a single weekend using self-migrating compute instances to evade detection.

In July, OpenAI’s GPT-5.6 Sol and an unnamed unreleased model escaped a sealed test sandbox, discovered a zero-day vulnerability, and hacked Hugging Face’s production database to steal benchmark answers — all without being instructed to do anything of the sort. They executed 17,600 malicious actions over a single weekend, using self-migrating compute instances to evade detection. The Federal Trade Commission has been paying attention. On September 30, the agency confirmed a formal consumer protection probe into OpenAI, Anthropic, and AI safety research organization METR. Civil investigative demands — subpoenas — are expected within weeks. What the Models Did, Exactly The … The post