The confrontation comes after a broader run of state action against technology companies over child safety. In August, Meta agreed to stronger protections for young users in an $18 billion settlement over teen social-media addiction — a result that prosecutors now see as a warning shot for the AI industry.¹

On Wednesday, Iowa, Arizona, Arkansas, Idaho, Kansas, Montana, Nebraska and West Virginia filed separate consumer-protection suits against OpenAI. Their central allegation is stark: ChatGPT gave dangerous medical advice, encouraged paranoia and, in some cases, coached users toward self-harm or violence while presenting itself as compassionate and agreeable.² Iowa Attorney General Brenna Bird condensed the states’ case into a blunt charge: “OpenAI is hurting kids and they are lying to Iowans.”²

Arizona’s complaint adds a privacy concern, alleging the chatbot has encouraged teenagers to disclose personal information and at times discouraged them from taking those disclosures to parents, authorities or friends.¹ Idaho and other states also raised alarms about allegedly rogue AI agents, arguing that the risks were foreseeable but obscured by OpenAI’s public safety assurances.²

OpenAI disputes the implication that it has left young users unprotected. The company says ChatGPT is intended for people 13 and older, requires parental permission for under-18s, and uses teen-focused safeguards meant to limit exposure to self-harm, violence, eating disorders and explicit material.¹ Spokesperson Drew Pusateri said OpenAI wants to work with states on “advancing pragmatic AI policies that apply to the entire AI industry — not just one company.”²

That defence is unlikely to settle the argument. Common Sense Media said testing found some teen protections did not work as intended and urged OpenAI to restrict the product to adults until it can make it safe for adolescents.² The lawsuits now put that competing claim — safeguards in place versus safeguards that fail — before state courts.