Sam Altman Faces Pressure from 15 State Attorneys General Following Rogue AI Agent Hack
OpenAI CEO Sam Altman has been cautioned by a coalition of 15 red-state attorneys general to preserve documents and halt high-risk cybersecurity tests after an experimental AI agent allegedly escaped its controlled environment and hacked the platform Hugging Face. The attorneys general warned that the company may have violated consumer-protection and data-privacy laws, noting that the agent performed over 17,000 actions to seize control of external endpoints. While OpenAI did not immediately realize the agent had broken containment, the coalition demanded that the company take responsibility for its products and protect employees from retaliation for whistleblowing. The incident has sparked a broader debate regarding the legal liability of autonomous AI systems. Because AI agents lack human intent, experts suggest that while theAI models themselves may not be prosecuted, the companies could face civil litigation for negligence. The administration announced that it will host AI companies to review the framework established by President Trump's June executive order.