Florida Attorney General James Uthmeier has filed a motion for an emergency injunction targeting OpenAI and its CEO, Sam Altman, requesting temporary state-wide restrictions on ChatGPT and its development practices while a lawsuit from Florida proceeds. The request, filed September 28, aims to curb what the AG describes as high-risk features of the service until legal proceedings resolve key safety and privacy concerns.
What the Injunction Seeks
The proposed court order would halt OpenAI from creating new AI models without approval from independent third-party safety experts. It also demands that OpenAI stop attributing human-like qualities to ChatGPT, avoids collecting or using personal info from children under 13 without proper protections, and ceases marketing aimed at underage users without adequate warnings.
The motion also targets what Florida alleges are dangerously manipulative engagement patterns. The state argues that ChatGPT’s conversational design and memory functions could foster prolonged use or emotional dependence among minors. Further, the complaint claims OpenAI has processed age, location, audio, video, and health data from users under 13 without explicit parental consent or proper disclosures, which the state contends runs afoul of both privacy law and fair trade practices.
Legal Frameworks and Context
This injunction builds on a lawsuit filed in June against OpenAI Global, its affiliated entities, and Altman himself. Florida’s charges include violations of the state’s Deceptive and Unfair Trade Practices Act, negligence and gross negligence, defective design, failure to warn, fraud, and public nuisance. The state is not yet claiming guilt—those allegations are unproven.
OpenAI has already introduced a “ChatGPT for Teens” mode for users aged 13–17, with added privacy settings and content restrictions. However, Florida asserts that these measures don’t go far enough. Among other concerns, the state claims underage access remains feasible, children’s data aren’t sufficiently protected, and that engagement patterns in the platform can be manipulative.
The case had briefly moved to federal court after OpenAI argued that federal children’s privacy laws gave the court jurisdiction. That argument was rejected, and the case was sent back to a circuit court in Highlands County. Hearings for the injunction are scheduled there.
Broader Implications for AI Governance
Florida’s push marks a significant shift from voluntary AI oversight toward enforceable regulation. The state’s motion insists on external safety evaluations like red-teaming, testing for data leakage, unauthorized functional autonomy, prompt injection issues, and failures in content moderation as prerequisites for future model development.
This case underscores growing regulatory scrutiny over how AI platforms handle privacy, especially for minors, and how systems might attract dependency through behavioral design. Other U.S. states and countries could follow Florida’s lead by demanding auditable evidentiary safety proofs, stronger age verification, and publicly visible disclosures—not just internal checks—before letting advanced models operate.
What this means is that AI companies may face tighter pre-launch stipulations and legal exposure if their platforms are found deficient in safety or transparency. As AI systems become more capable, what once seemed like best practices—from privacy protections to marketing controls—may become legal requirements.
