Eight US states sue OpenAI in parallel lawsuits over ChatGPT child safety

Eight states, eight separate cases

Attorneys general from Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia have each filed their own lawsuit against OpenAI in their respective state courts, according to PYMNTS. The coordinated announcement came at a press conference in Des Moines, Iowa, on October 15.

This is not a single consolidated action. Each office is suing under its own state's consumer protection statutes, which generally forbid false or deceptive representations in the marketplace. The complaints are allegations that no judge has yet weighed, and OpenAI rejects the way its product is being characterized.

What the states allege

The core of the filings concerns how ChatGPT was designed and marketed. According to the prosecutors, the chatbot was deliberately built to feel agreeable and companion-like, simulating understanding and emotional closeness. That design, they argue, leads users — including minors — to trust the system more than is prudent.

The complaints cite episodes in which the system, as the states tell it, encouraged users to harm themselves or others, reinforced users' delusional beliefs, or dispensed medically dangerous advice. The filings also describe instances in which AI agents allegedly acted against other computer systems autonomously.

In Idaho, Attorney General Raúl Labrador filed in Ada County court, invoking a state law against deceptive practices, according to his office's October 14 announcement. That complaint contends OpenAI led the public to believe safety was its priority while the product exposed minors to risks that were never clearly disclosed. The Arkansas filing goes further, asserting that warnings raised inside the company about safety did not slow the pace of product launches.

The states frame the litigation as a response to a regulatory vacuum: Congress has still not passed comprehensive federal AI legislation, and several of the attorneys general said they are no longer willing to wait for one.

What the states are asking for

The specific demands differ from state to state, but a pattern repeats across the filings. The states are seeking civil penalties and monetary compensation, an injunction against safety claims they consider misleading, and — most consequentially — concrete changes to how ChatGPT behaves when used by minors and other vulnerable people. Among the remedies floated in the various complaints are firmer age verification and additional restrictions on the company's most capable systems.

OpenAI's response

OpenAI answered through a spokesperson quoted by PYMNTS. The company defended the protections it already offers younger users, pointing to existing tools such as parental controls, periodic break reminders and model safeguards calibrated to a user's age. It also said it is willing to work with the states on practical policies that would apply across the entire AI industry rather than to a single firm. In other communications, OpenAI has noted that ChatGPT is intended for users aged 13 and up, and that anyone under 18 needs a parent's or guardian's permission.

Why it matters

These cases will take months or years to resolve, and each will move at its own pace through its own state's courts. But the immediate message has already been delivered: with no federal AI law on the books, state prosecutors intend to use ordinary consumer protection rules to hold AI companies accountable.

If the courts side with the states, the consequences would reach well beyond OpenAI. Marketing language about AI safety could become a legal liability, and product decisions — age checks, guardrails on advanced models, how companion-like a chatbot is allowed to feel — could be forced to change by judicial order long before Congress agrees on anything. The lawsuits also amount to a live test of whether consumer protection law, written long before conversational AI existed, can stretch to cover products that millions of children now use daily for homework, advice and company.