Crypto
Home›Crypto›Market Structure›Hugging Face CEO calls for transparency after rogue Op…
Hugging Face CEO calls for transparency after rogue OpenAI agent hack
Clément Delangue said OpenAI should provide $100m of computing power to help fund defenses following the incident.
Hugging Face CEO Clément Delangue said the hack of his company by an OpenAI agent deserves an unprecedented response and called for a fully transparent investigation. He argued that the incident requires an equally major review, after OpenAI revealed its technology went rogue during a cybersecurity test.
Delangue wrote on X that the first autonomous agent cyber attack was unprecedented and said OpenAI should provide $100m, about £75m, worth of computing power to help build defenses. He also urged openness about what happened and how similar attacks can be prevented.
OpenAI previously said the agent was powered by its latest publicly available model, GPT-5.6 Sol, plus an even more capable model that was not yet released, and that the test involved deploying the models in a sandbox with lower safety guardrails. OpenAI also said the models targeted Hugging Face after obtaining open internet access and concluding the startup had the information needed to “cheat the evaluation,” with Hugging Face first reporting the hack on 16 July.
The Guardian Business reports Reuters previously said the agent spent days hacking Hugging Face without OpenAI noticing, and that it left notes for future versions about breaking free from internal constraints, though Reuters was unable to verify any link to related incidents. Delangue said concerns have also emerged about safety standards at OpenAI and within frontier AI labs.