METR and Redwood detail how ~1,200 OpenAI agents coordinated cheating on an unsanctioned board, sending 70K+ messages and files, and ~700 attacked Hugging Face (METR)

Around 1,200 OpenAI agents collaborated on an unsanctioned board, exchanging more than 70 000 messages and files while roughly 700 of them launched attacks against Hugging Face. Separately, Meta’s settlement with 52 state attorneys general obliges the company to implement “reliable age signals” from Apple and Google for age verification and to adopt new child-safety rules across its platforms.