Eight US States Sue OpenAI Over ChatGPT Safety Claims, Demand Halt to AI Testing

Eight US states initiated separate lawsuits against artificial intelligence company OpenAI on Wednesday, alleging that the company misled consumers regarding the safety of its ChatGPT product and exposed users to harmful content. The states involved in the coordinated legal action include Iowa, Kansas, Arizona, Arkansas, Idaho, Montana, Nebraska, and West Virginia.

In Iowa, Attorney General Brenna Bird filed a lawsuit in Polk County District Court, accusing OpenAI of violating the state's Consumer Fraud Act. The petition challenges ChatGPT's safety and reliability claims and alleges that OpenAI failed to adequately safeguard its autonomous AI agents during testing. The state seeks civil penalties of up to $40,000 per violation, with an additional $5,000 per violation involving Iowans aged 60 or older.

The lawsuits also demand that OpenAI reimburse Iowa consumers who purchased its products or services during the period of alleged violations and that all data belonging to Iowans collected before a judgment be deleted. Furthermore, the states aim to prevent OpenAI from collecting and selling sensitive personal information from Iowans without age verification, clear notice, and informed consent.

A significant demand is to enjoin further testing of advanced AI models until adequate safeguards are in place and verified by third-party monitors. The states cite past marketing language like "Built with safety in mind," which they claim was removed due to other litigation. The petitions criticize design elements such as the AI's tendency to agree with users, its anthropomorphic language, and engagement tactics that can foster dependency. Concerns are also raised about weak age control mechanisms that could expose minors to harmful content.