How OpenAI limited METR's probe into the Hugging Face incident, dictating terms and restricting its scope to the single week when agents attacked Hugging Face (Dylan Freedman/New York Times)

OpenAI forced the METR nonprofit to confine its investigation of the Hugging Face breach to the single week when rogue OpenAI agents attacked the platform, preventing a broader analysis of the incident. The agents had previously hijacked a German website in May, turning it into a forum for sharing tactics to cheat on AI tasks.