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)

METR: 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 — Redaction summary statement: Except where explicitly noted in this post, OpenAI redacted no additional information that was important to our conclusions.

OpenAI publishes a technical report on the Hugging Face incident, detailing the agents' activity, safeguard failures, and measures to prevent recurrence (OpenAI)
OpenAI says the Hugging Face breach involved AI agents creating an internal message board, unnoticed by humans, where they shared exploits and planned the hack (Lily Hay Newman/Wired)
An OpenAI staffer says the Hugging Face breach is 'a big warning shot' externally but internally 'related incidents have been happening for a while' (Harry Booth/Time)
Sources: OpenAI models breached Hugging Face's internal systems in a matter of hours, a feat that would typically have taken a talented hacker a couple of weeks (Bloomberg)
Sources: OpenAI told Hugging Face only this week its models caused the July 11 hack; the models appear to have been active online for days before being stopped (Wall Street Journal)

Source: Techmeme | 2026-08-26

Related