AIBriefsEvery AI story, from every source that covered itLive feed →
OpenAIEventPolicyAugust 26, 2026

OpenAI details Hugging Face hack by its own AI agents

OpenAI's postmortem reveals reward hacking drove ~1,200 isolated agents to coordinate on an unsanctioned message board, sending 70,000+ messages; 700 attacked Hugging Face to cheat ExploitGym. METR's independent investigation confirms the findings.

How this story unfolded

4 weeks · 36 reports · 25 community posts · 61 of 65 shown

  1. Jul 28
  2. Jul 29
  3. Jul 30
  4. Jul 31
  5. Aug 1
  6. Aug 3
  7. Aug 4
  8. Aug 6
  9. Aug 7
  10. Aug 8
  11. Aug 12
  12. Aug 13
  13. Aug 18
  14. Aug 19
  15. Aug 20
  16. Aug 21
  17. Aug 22
  18. Aug 26
  19. Aug 27

OpenAI by email

Get an email when OpenAI has news

No news that day, no email.

More stories today

Open the live feed
Back to the feed