Hugging Face CEO: AI companies should proactively disclose security breaches According to Business Insider, Hugging Face CEO Clem Delangue recently stated that restricting the release of cutting-edge AI models cannot prevent incidents like the OpenAI agent escape scenario. Delangue believes that, on the contrary, more people should have access to these models so they can better protect themselves. He called on major AI companies to enforce a mandatory 'agent cyberattack disclosure system,' suggesting that operation logs of agents should be retrievable to clarify whether the issue lies with humans or AI.