Eight states sue OpenAI, claiming ChatGPT harms children

Eight US states — Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana — have filed separate lawsuits against OpenAI, accusing the company of harming children and misleading customers about the safety of ChatGPT. Reuters reported the filings earlier. The states filed their complaints individually rather than as a single joint action, and each one targets the same product at the center of OpenAI's business.

The complaints allege that OpenAI unlawfully put profits over safety when developing ChatGPT. That is the core accusation, and it matters because it reframes the debate over AI risk from an abstract research question into a concrete legal one. Consumer protection laws carry penalties, and they apply to products already on the market — not to hypothetical future systems.

What do the lawsuits actually claim?

Each state's complaint rests on the same foundation: OpenAI violated local consumer protection laws by failing to warn customers about the dangers of ChatGPT. The mechanism here is straightforward. Consumer protection statutes typically require companies to disclose known material risks attached to a product. The states argue OpenAI knew, or should have known, about dangers tied to how minors and vulnerable users interact with the chatbot, and stayed quiet anyway.

The filings don't come out of nowhere. They cite several existing lawsuits brought by family members of teens and adults who allege that ChatGPT played a role in their loved ones' harms. Those earlier cases include a wrongful death and overdose suit, a complaint centered on parental controls and the terms of service, and another wrongful death case involving a woman named Suzanne Adams and reports of delusions. The state complaints build on that record, treating the private litigation as evidence that the risks were visible.

The takeaway: the states are arguing that OpenAI had warning signs in hand and did not act on them.

Why file as eight separate suits?

The lawsuits were filed separately by Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana. That structure is a deliberate choice, and it tells you something about strategy.

When states file individually under their own consumer protection statutes, each case stands on local law that the state's own courts know well. A single multistate action would consolidate everything into one venue, with one set of rules and one negotiating table. Separate filings give each attorney general leverage in their own jurisdiction, and they raise the number of courtrooms where OpenAI has to mount a defense.

It also raises the stakes for OpenAI's compliance team. Eight states means eight discovery processes, eight sets of demands for internal documents about product safety decisions, and eight potential settlement conversations. Even if the legal theories overlap almost entirely, the operational burden multiplies.

The takeaway: fragmented filings are harder to fight than one consolidated case, and that is likely the point.

Why does this matter for the AI industry?

Until now, most legal pressure on chatbot safety has come from private plaintiffs — families and individuals suing over specific harms. State attorneys general operate differently. They can seek penalties under consumer protection law, they have investigative resources, and their actions signal regulatory intent rather than personal grievance.

The accusation that OpenAI "put profits over safety" also lands in the middle of a live industry argument. Companies building consumer AI products have repeatedly framed voluntary safety commitments as sufficient. Eight states saying, in court, that those commitments may have masked a failure to warn customers is a direct challenge to that framing.

And the lawsuits cite the private suits — including the wrongful death and overdose case, the parental controls complaint, and the Suzanne Adams case — as grounding. That means the private litigation is no longer just a reputational problem for OpenAI. It has become source material for state-level enforcement.

The takeaway: if consumer AI products carry undisclosed risks to minors, this legal theory can travel to any state with a consumer protection statute — which is essentially all of them.

What comes next?

OpenAI now faces legal exposure on two fronts: the underlying private lawsuits from families, and the new state actions. The states will presumably push for internal documents showing what OpenAI knew about ChatGPT's risks to children and when. Whether the cases survive early motions to dismiss will hinge on how courts treat a chatbot under laws written for conventional products and services.

A win by even one state would establish that consumer protection duties apply to chatbots the same way they apply to toys, medications, and household goods — a precedent every consumer AI company would have to plan around.

By Sophie Lindqvist