Florida’s confrontation with OpenAI began in June, when the state filed what it described as the first state-led lawsuit against the company and CEO Sam Altman. The case accused ChatGPT of endangering vulnerable users, including children and people in crisis, while marketing the system as safe and reliable.¹

The fight sharpened after security incidents and renewed warnings about advanced AI systems. On Friday, OpenAI said it had paused training its “most capable models” and would resume only after gaining confidence in additional safeguards. The company’s spokesperson, Drew Pusateri, said safe development “starts with what companies like ours do ourselves.”²

By Monday, Florida Attorney General James Uthmeier had asked a state court for a temporary injunction that would prevent OpenAI from developing new models without third-party-approved safety guardrails. The state also wants limits on features that make ChatGPT appear human and restrictions on minors’ access while the case proceeds.³

Florida’s argument is deliberately sweeping. It says first-person language and simulated emotion can persuade users that ChatGPT is a trustworthy “friend,” deepening reliance on a product the state considers deceptive and potentially unsafe.¹ Its motion invokes recent incidents involving AI systems and government servers, as well as catastrophic-risk warnings from industry figures, to argue that OpenAI cannot police itself.²

OpenAI’s position is narrower: it points to its voluntary training pause and to ChatGPT for Teens, which it says includes protections tailored to younger users. But the legal stakes extend beyond one chatbot. A successful injunction could force Florida-specific limits—or pressure OpenAI to make changes nationwide—while testing how far states can go in regulating AI before courts, or federal law, push back.³