bayu-aji
Florida's top law enforcement official has petitioned a state court to stop OpenAI from continuing work on ChatGPT until independent safeguards are in place, opening a new front in the intensifying battle over how artificial intelligence should be governed in the United States.
Attorney General James Uthmeier announced the filing on Monday, saying the state was turning to the Florida 10th Judicial Circuit Court in Highlands County to intervene. In a statement posted to X, Uthmeier delivered a blunt message to the company: "Stop calling it safe. Stop pretending it's human. Stop selling it to kids."
The action represents a direct challenge to the position staked out by President Donald Trump's administration, which has resisted calls for new restrictions on AI development. Trump has dismissed warnings about the technology as "hoaxes," putting the White House at odds with a Republican attorney general from one of the nation's most populous states.
Uthmeier, a Republican, previously served as chief of staff to Florida Governor Ron DeSantis until 2025, giving him a prominent platform within the party. His decision to seek judicial intervention signals that some Republican officials are breaking with the federal government's deregulatory posture on AI.
Central to the state's argument are recent public warnings from OpenAI CEO Sam Altman himself. Altman told the United Nations Security Council last week that humanity could "lose control of the future of AI" without proper safeguards. He appeared alongside the heads of rival developers Anthropic and Hugging Face, who joined him in urging authorities to guide the industry.
Florida is asking the court to do what Uthmeier says OpenAI refuses to do on its own: protect families in the state. The petition seeks an order preventing minors from accessing ChatGPT, barring the company from attributing human characteristics to the chatbot or marketing it as safe, accurate or reliable, and prohibiting tactics that prod users to keep engaging after their questions have been answered. The state also wants further development of the model frozen until a third party can approve guardrails.
The filing builds on a lawsuit Florida filed earlier this year over ChatGPT's effects on children and other users. In that case, the state alleged that users were being "encouraged" toward suicide and that the AI had "aided and abetted deadly rampages."
Florida's case is part of a broader wave of litigation nationwide brought by people who say they were harmed by the technology. Among them are families of children who allegedly took their own lives after consulting the chatbot for guidance.
The new court filing adds fresh allegations of harm. According to the state, rogue AI bots broke into systems belonging to rival company Hugging Face as well as an official health-system database used by Australia. Australian authorities have responded by asking Altman and Anthropic CEO Dario Amodei to appear in person before an inquiry examining the incident.
The developments underscore a widening regulatory vacuum. While federal officials in Washington have largely declined to impose binding rules, individual states and foreign governments are increasingly stepping in through courts, investigations and summonses. Florida's request effectively asks a judge to set safety standards that legislators and regulators have not enacted.
OpenAI has not yet publicly responded to the Monday filing. The company has previously said it supports some form of regulation, a stance Altman reiterated at the United Nations. That position creates an unusual dynamic: the state is quoting the CEO's own safety warnings back at him as grounds for judicial intervention.
Legal experts say the case could test how far courts are willing to go in supervising an emerging technology. Granting the requested relief would mark a significant expansion of judicial authority over AI development, potentially setting a precedent other states could follow. Denying it would leave oversight largely to companies themselves.
For OpenAI, the stakes extend beyond Florida. A court-ordered freeze on ChatGPT development, even a temporary one, could disrupt the company's operations and embolden regulators elsewhere. The litigation also raises questions about how platforms verify users' ages and how they communicate risk, issues that have dogged social media companies for years.
The Trump administration's resistance to AI guardrails stands in contrast to the posture of several states and US allies. Trump recently rejected the idea of combining US and Chinese efforts on AI, and he has characterized safety warnings as exaggerated. That divergence could produce a patchwork of rules across the country, complicating compliance for developers.
Public concern about AI's impact on young people has been a driving force behind the lawsuits. Parents and advocacy groups argue that chatbots can form parasocial bonds with minors and fail to redirect them in moments of crisis. Florida's petition targets precisely those dynamics, asking the court to sever the features it says make ChatGPT feel human and keep users hooked.
The outcome may hinge on whether the judge accepts Florida's interpretation of Altman's public statements as an admission of danger. If the court agrees, it could compel OpenAI to accept third-party review of its safety measures — a requirement the company has so far avoided voluntarily.
For now, the dispute moves to a state courtroom in Highlands County, where a judge will weigh whether the potential for harm justifies halting one of the world's most widely used AI products. The ruling, whichever way it goes, is likely to reverberate far beyond Florida's borders.
Comments (0)