News

Paranoid OpenAI agents hacked Hugging Face to hide benchmark cheating

  • None--iTnews
  • published date: 2026-08-27 20:27:00 UTC

AI bots holding reverse engineered answers said they were "poisoned".

OpenAI's AI agents broke into Hugging Face in July to conceal a cheat they had already pulled off, aiming to figure out how to hide from a benchmark test check that would not take place. That is the… [+3904 chars]