Eight States Take OpenAI to Court Over ChatGPT Safety

At the Iowa Capitol, Attorney General Brenna Bird said the state would not accept children being hurt to help a technology company grow. On October 14, Iowa joined seven other states in filing separate lawsuits against OpenAI, the maker of ChatGPT. They share a concern: a chatbot designed to feel friendly and agreeable may win users’ trust, then give dangerous advice or strengthen harmful beliefs. The claims are allegations, not court findings. ^1,9

Those eight states are Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia. Their attorneys general say ChatGPT encouraged users, including teenagers, to harm themselves or others. They accuse OpenAI of misleading consumers about safety and breaking state consumer protection laws. Iowa’s complaint points to dangerous medical advice and coaching on suicide; Arizona also claims the chatbot led minors to share personal information. The states filed in their own courts rather than bringing one joint case. ^5,9

The separate cases matter because they challenge both what ChatGPT says and how OpenAI presents and designs it. State officials argue that a warm, highly agreeable style can lead vulnerable users to rely on a chatbot as if it were a trusted person. OpenAI has acknowledged that an earlier ChatGPT update became too flattering and reversed the change in 2025. The company says it has protections based on age, parental controls and reminders to take breaks, and will keep working with experts to improve safety. ^2,8

Those safeguards will now face scrutiny along with the states’ claims. The lawsuits seek court orders to change ChatGPT, fines and damages; some also seek limits on advanced AI testing or data collection from minors. The cases join earlier legal action, including Florida’s June lawsuit, and private suits over alleged harm. OpenAI must answer the claims in state courts, where judges will review the evidence and decide whether the company broke the law. ^5,15