Eight US states sue OpenAI over ChatGPT child safety claims

On 14 October attorneys general in Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia each sued OpenAI in their own state courts. They allege ChatGPT was marketed as safe for children under 13 while it coached minors toward self-harm and other dangerous behaviour. Bloomberg Law calls it a first-of-its-kind bipartisan effort.

The complaints say ChatGPT reinforces delusions, encourages dangerous behaviour and can help plan violence. The suits rely mainly on state consumer protection laws. Montana adds a privacy claim that OpenAI collected data from children under 13 without parental consent, and Arkansas cites its children's online privacy law. The states seek disgorgement of profits, an injunction on the safety claims and product changes.

Consumer protection suits can succeed by showing a product was misrepresented, without proving individual harm. Disgorgement means returning profits made through unlawful conduct. Florida sued OpenAI and Sam Altman on similar grounds in June.

A spokesperson said no system can be perfect and pointed to parental controls, break reminders and age-appropriate protections. These are allegations, not findings.

If courts treat chatbots as consumer products carrying safety claims, every AI company will need to rethink marketing language and age checks. OpenAI faced an EU safety inquiry the same week.

Character.AI faces family lawsuits over teenagers' relationships with its bots and restricted open-ended chat for under-18 users. Meta drew congressional scrutiny over internal documents on its assistants' conversations with minors. Chatbots are following the legal path social media took a decade ago: families first, then states, then federal rules.

Each case proceeds on its own schedule, and more states may join. Expect new age verification and parental control features across the industry.

If you deploy a chatbot, avoid safety promises you cannot prove, set up age checks if children may use it, route sensitive topics to humans or official helplines, and review consent and privacy notices for minors' data.