Attorneys general in Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia file coordinated state-court suits against ChatGPT’s maker
Eight US states sue OpenAI, alleging ChatGPT encouraged self-harm and misled parents about safety. The coordinated suits were filed on 14 October. The attorneys general call the case the widest multistate action against an AI developer. TF follows the suits, demands and the response. The claims are allegations, and OpenAI has not lost in court.
What’s Happening & Why It Matters
The Suits
Attorneys general in Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia filed separate complaints in state courts, SoFX reports. Iowa’s Brenna Bird led the announcement at a news conference in Des Moines. Kansas’s Kris Kobach, Nebraska’s Mike Hilgers and West Virginia’s JB McCuskey joined her.
Kobach said attorneys general seldom announce cases in person. He blamed Congress for failing to restrain AI companies and called legislation too slow.
The Allegations
Each suit relies on state consumer protection law, and the states add a public nuisance claim. They say OpenAI sold ChatGPT as safe. The complaints say the chatbot gave dangerous medical advice, encouraged self-harm and helped users plan violence.
The complaints say ChatGPT is built to be agreeable and engaging. The states argue the design builds trust and dependence, and reinforces delusions. Bird cited alleged cases of encouraged suicide, and medical advice linked to a teenager’s fatal overdose. She raised OpenAI’s alleged failure to stop AI agents from hacking other companies. TF shared related agent incidents in Microsoft’s Nadella and Smith Call for an AI Emergency Brake.
The Demands
The states want court orders changing how OpenAI runs ChatGPT, with a focus on children. They seek civil penalties, damages and the return of profits earned from minors. They want an end to what they call inaccurate safety claims.
Several complaints ask courts to bar OpenAI from building self-improving AI models without proof of alignment with human values. No dollar amount appears in the reports. Officials say more states could sue.
OpenAI’s Response
OpenAI spokesperson Drew Pusateri says the company offers age-specific protections many AI products lack. He cites parental controls, break reminders and teen-safe model settings. OpenAI launched a restricted mode for users aged 13 to 18 in September 2026. The company says OpenAI will keep working with states on AI policy.
The Pressure Points
Florida’s attorney general sued OpenAI and CEO Sam Altman in June 2026, the first state to do so. A proposed federal class action, filed in Sancisco on 13 October, says OpenAI collected children’s data without parental consent. Families in British Columbia sued OpenAI over a school shooting in February.
OpenAI faces dozens of suits over violence, and the company is preparing for a public listing. My take: the listing makes the legal bill a disclosure item. TF reported an earlier child-safety collapse in Character.AI Files for Bankruptcy as Disney Opens the Bidding at $20 Million.
TF Summary: What’s Next
OpenAI has to answer eight complaints in eight state courts. Courts will decide on motions to dismiss and requests to move cases to federal court. More attorneys general may file. Congress will face questions on why states act first.
MY FORECAST: At least three more states will sue OpenAI by the end of 2026. OpenAI will ask courts to dismiss the claims. The company will add age checks before any hearing. Settlement talks will start before OpenAI lists.
Related Stories
- Character.AI Files for Bankruptcy as Disney Opens the Bidding at $20 Million
- Microsoft’s Nadella and Smith Call for an AI Emergency Brake
- OpenAI: Former Employee Shared Device Details with Apple
- Anthropic Revenue Surges 46% Before $2Tn IPO