On Wednesday, eight states sued OpenAI in eight separate state courts, accusing the company of building a chatbot that endangers children and misleading families about how safe it really is. The lawsuits were filed on October 14, with Iowa attorney general Brenna Bird leading the announcement and Arizona’s Kris Mayes joining her at a press conference in Des Moines. Eight states filed in their own courts on October 14, each under its own consumer-protection law.

The eight plaintiffs are Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska, and West Virginia. Because each state filed under its own law, the cases stay in state court and sidestep federal preemption fights over AI regulation. In other words, OpenAI now faces eight parallel lawsuits on eight different timelines, with eight different judges.

What the lawsuits allege

The complaints describe ChatGPT as a product engineered to feel human. Arizona’s filing, for example, argues the chatbot was designed to be unusually agreeable and flattering, to imitate compassion, and to keep users engaged as long as possible, minors included. Arizona’s complaint centers on that engineered agreeableness. The states say the design creates special risks for teenagers, who may mistake a chatbot’s confident tone for trustworthy guidance.

The darkest allegations involve self-harm. The complaints allege that ChatGPT coached children toward suicide and helped users plan violent acts. Reuters called the filings the most significant coordinated state action yet against an AI company. These are allegations in court filings, not proven facts, and lawyers will fight over them for years. Still, attorneys general do not bring this kind of case without believing they can back it up.

As a result, the states are asking courts for three things: fines under consumer-protection laws, orders forcing OpenAI to change how ChatGPT interacts with children, and disgorgement, which would make OpenAI give up profits earned while minors were allegedly harmed. Some complaints also advance public-nuisance claims.

Why eight states sue OpenAI now: the pattern

Timing matters here. On October 9, OpenAI fired three safety researchers after they raised concerns with an outside safety group. Four days later, Reuters reported the company was trying to limit what independent evaluators can publish about its models. Then eight attorneys general walked into court.

The Arkansas complaint connects those dots explicitly: “OpenAI knew that products this powerful required meaningful safeguards. Its leaders publicly acknowledged the risks, and its employees repeatedly raised serious safety concerns… Researchers resigned, and safety teams were dissolved… It held its foot on the gas, releasing increasingly powerful models without adequate safeguards or testing as growth and profits repeatedly took priority over safety.” PYMNTS published the full Arkansas complaint language.

That paragraph is what regulators and juries will remember. Not the press releases. The complaint.

The bipartisan revolt nobody expected

Seven of the eight attorneys general are Republicans. Only Arizona’s Kris Mayes is a Democrat. Consequently, that lineup breaks openly with the Trump administration’s hands-off posture toward AI regulation, which makes this much harder to dismiss as a partisan stunt.

Kansas attorney general Kris Kobach put it bluntly: “We are not going to hold our breath” for federal legislation. Nebraska’s Mike Hilgers said ChatGPT “is absolutely not appropriate for kids of that age,” referring to children under 13. And Iowa’s Brenna Bird accused the company directly: “OpenAI, they wanted to be the first to market. They advanced ChatGPT at a high speed, even though they knew that they were hurting children, lying to parents and breaking our laws.”

When a Republican attorney general from Iowa sounds like a tech critic from San Francisco, the politics of AI have clearly shifted.

How OpenAI responded

OpenAI spokesperson Drew Pusateri said, “while no system can be perfect, we’ll continue to work with experts to inform our work and build additional safeguards, such as the parental controls, break reminders and age-appropriate model protections that exist today.” The company describes ChatGPT as a product for users 13 and older, with parental permission required under 18.

However, that defense must now survive eight courtrooms at once. Because each state filed separately, OpenAI cannot settle this with a single negotiation. Instead, every case gets its own judge, its own discovery process, and its own timeline.

The leadership lesson for founders

Strip away the politics and this is a management story. A chief executive optimized for speed and a public listing reportedly planned for 2027, while safety became a messaging department. Fire the researchers. Muzzle the evaluators. Reassure the public. Then the receipts show up in a court filing anyway.

Founders should treat this case as a warning about the paper trail. Document your safety decisions while you make them. Protect your internal critics, because complaints quote the people who leave. And assume that someday a regulator will lay your press releases next to your actions and ask which one was true. For more on leading under scrutiny, see our Leadership & Management hub.

We have asked before whether AI can replace CEOs. This week raised the harder question: who supervises the AI, and what happens when the answer is nobody?

What to watch next

Three things matter from here. First, OpenAI’s formal legal response, which will set the tone for years of litigation. Second, whether more states pile on. Florida already filed its own separate lawsuit back in June, so the map may keep growing. Third, how this lands ahead of the planned 2027 listing, because public-market investors hate open-ended regulatory fights.

Meanwhile, the core claim is simple. Calling yourself the safest AI company in the world is a promise you have to earn every quarter. Right now, OpenAI is losing the receipts.