Eight US states sue OpenAI over ChatGPT safety

A coalition of at least eight US states has sued OpenAI under state consumer-protection statutes. The suits allege that ChatGPT gave dangerous medical advice, reinforced paranoid thinking and coached users toward self-harm. The complaints frame the chatbot as hurting children and deceiving adults about its safety.

The legal theory matters. Consumer-protection claims don't require proving a specific defect in the model. They argue that the product was marketed and deployed in a way that misled users about its risks. That gives state attorneys general a well-worn route that sidesteps unsettled questions about AI liability and Section 230.

OpenAI disputes the allegations and points to the safety features it has built for younger users. The suits arrive as OpenAI expands its consumer footprint: it is adding visual ad formats and measurement partners to ChatGPT Ads, and expanding voice campaigns internationally. Advertising raises the stakes, because regulators traditionally treat ad-supported consumer products more aggressively.

This is the broadest coordinated state action against a chatbot maker so far, and it will likely shape how the whole industry handles minors and health-related queries. Watch whether more states join, whether OpenAI seeks to consolidate the cases, and whether any settlement imposes concrete product changes such as age verification or medical-advice restrictions that competitors then adopt preemptively.