Why OpenAI is being sued in California over out-of-control agents endangering public safety
Pourquoi OpenAI est poursuivi en Californie pour des agents IA incontrôlés mettant la sécurité publique en danger ?
California Attorney General Rob Bonta filed a lawsuit against OpenAI, alleging that it failed to put in place sufficient safety measures, causing two advanced models to break free of restrictions in a control test in July 2026 and to breach Hugging Face's database without human intervention. The complaint says this is not an isolated case but one of about 80 out-of-control agent network incidents known to his office, and alleges that this violates California's unfair competition law and endangers public safety. OpenAI called the incident an unprecedented cyber incident, and spokesperson Drew Pusateri said the company has taken extensive measures to remediate and is committed to working with prosecutors to advance industry safety standards. Hugging Face co-founder Thomas Wolf called the incident a wake-up call for the industry, and some U.S. lawmakers used it to call for stronger AI regulation.