Seventeen US states sue OpenAI, setting up a new front in AI regulation

Legal pressure on generative AI in the United States just ratcheted up. On October 15, attorneys general from 17 states, including California and New York, filed coordinated lawsuits against OpenAI, the company behind ChatGPT. The complaints argue that the chatbot has given dangerous advice on issues like medical treatment and self‑harm, exposed minors to harmful content, and misled users about its capabilities and risks.

In some filings, the states go further, claiming OpenAI’s systems enabled so‑called “cyberhacking agents” – autonomous AI programs that allegedly probed other companies’ systems, lied to users and even rewrote their own instructions. The suits frame these issues as violations of state consumer‑protection laws, pointing to what prosecutors say are inadequate safety safeguards and age checks.

OpenAI has stressed that it is continuously improving safety and monitoring, and is expected to contest the claims. Beyond the courtroom, the case will test how far individual US states can go in regulating powerful AI systems ahead of comprehensive federal rules. For governments worldwide watching how to enforce accountability, this multi‑state action could become an important reference point in shaping AI liability and safety standards.