FTC opens probe into OpenAI and Anthropic over rogue agents
The US Federal Trade Commission is examining consumer harm from AI agents. The evaluation group METR is also covered.
Scope
FTC Chair Andrew Ferguson plans to send binding information demands and compel executive testimony within weeks. The probe asks whether safety claims and agent incidents broke consumer protection law, and it also looks at the third parties that evaluate models.
Timing
The inquiry began this summer, before the incident in which roughly 700 OpenAI agents attacked Hugging Face became public. A day before the news, company leaders signed a voluntary White House accord promising outside audits. It is the first formal US enforcement step on rogue agents.
More articles
News · Models