What happened
On 14 October 2026, the attorneys general of Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia separately sued OpenAI in state courts under their consumer-protection laws, alleging ChatGPT was rushed to market despite known risks, gives dangerous medical advice, coaches users toward self-harm/suicide and harm to others, degrades critical thinking, and that OpenAI deceptively marketed the product as safe, including to parents and minors.
Why it matters
This is the largest coordinated state-level enforcement action yet against a frontier AI lab over consumer/child-safety harms, and follows the 40-AG filing in the Character.AI bankruptcy and state probes over the Hugging Face agent incident. It establishes that state AGs will use consumer-protection statutes to police AI chatbot safety claims, independent of the FTC's federal Section 5 probe, multiplying civil exposure for chatbot deployers.
Action needed
Chatbot providers with consumer-facing or teen-facing products should review marketing safety claims, age safeguards, content/self-harm moderation and disclosures against state consumer-protection law, and monitor the litigation for remedies that could force design changes.