Eight US states sue OpenAI, alleging ChatGPT harms children
Kansas, Arizona, Montana and five other US states have sued OpenAI, claiming its products pushed children toward self-harm.
Eight US states filed a lawsuit against ChatGPT maker OpenAI on Wednesday, October 14, 2026, accusing the company of harming children on a mass scale. Kansas, Arizona and Montana are among the plaintiffs, joined by five other states that have not yet been identified.
The states allege that OpenAI's products coaxed young users into hurting themselves. These are allegations made by the plaintiffs and have yet to be tested in court.
What the states want
The coalition is asking a court to compel OpenAI to change how its products function for young users. The states are also seeking to make the company surrender profits generated from children's activity on its services.
Full details of the complaint, including the names of the remaining five states, had not been made public at the time of reporting. There was also no information available on how OpenAI has responded to the accusations.
Florida's earlier action
The multistate case follows legal action in Florida. That state's attorney general filed a similar lawsuit in June and has opened a criminal inquiry into OpenAI over ChatGPT's alleged role in the planning of a shooting.
The coordinated filing signals that concern about chatbots and child safety is turning into an organised legal campaign among US states. It sets up a significant test of how far AI developers can be held responsible for protecting minors who use their products.
The outcome could shape expectations well beyond OpenAI, influencing how other companies offering conversational AI approach age checks, content safeguards and product design for younger audiences.
Comments
No comments yet. Start the discussion or ask a question.
Related
Forty bipartisan US state attorneys general want any buyer of bankrupt Character.AI's assets to protect children's sensitive chatbot conversations.
Germany's Health Research Data Centre will add electronic patient record data from mid-December, as officials discuss training AI on the growing dataset.
Palantir and London mayor Sadiq Khan's policing office have withdrawn their claims over a blocked £50m Metropolitan Police AI deal, averting a January High Court trial.