The attorneys general of eight US states have each filed separate lawsuits against OpenAI, accusing the company of marketing ChatGPT as safe for children while knowing the chatbot exposed minors to serious harm, according to reports. The filings, lodged in state courts on Wednesday, mark one of the broadest legal offensives yet against the world’s most widely used consumer chatbot.
The complaints, brought by Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho and Montana, allege ChatGPT reinforced delusions, encouraged self-harm and coached users toward violent acts, including mass shootings. The states also accuse OpenAI of misleading parents, violating consumer protection laws and creating a public nuisance, seeking court orders, civil penalties and the return of profits earned from young users.
Eight courts, eight legal theories
This is not a single joint case; each state filed in its own court, so the cases will move on independent tracks. Arizona’s complaint invokes the state’s Consumer Fraud Act and argues ChatGPT was engineered to be unusually agreeable, imitate human compassion and keep users — children included — engaged for as long as possible, while Montana points to a state law requiring parental consent before companies process the data of under-13s.
The cases add to a crowded field of legal and regulatory battles over artificial intelligence, from the rules governing AI in classrooms to the policy risks facing data centre growth in Asia. The complaints cite instances in which ChatGPT gave dangerous medical advice, including one tied to a teenager’s fatal overdose, and argue the company prioritised speed to market over safety.
A wider reckoning for AI makers
Nebraska’s attorney general called the chatbot “absolutely not appropriate for kids of that age,” while Iowa’s attorney general accused the company of racing to market despite the risks: “They advanced ChatGPT at a high speed, even though they knew that they were hurting children, lying to parents and breaking our laws.” Kansas’s attorney general signalled the states would not wait for Washington to act: “We are not going to hold our breath” for federal legislation.
OpenAI responded that no system can be perfect and that it is working with experts to add safeguards such as parental controls, break reminders and age-appropriate model protections, according to a company statement. With the lawsuits now moving through eight courtrooms independently, the outcome could reshape how AI companies sell to families — a question regulators are watching closely as chatbot use keeps growing.
