Florida AG asks court to halt OpenAI model training without oversight

Florida Attorney General James Uthmeier asked a state court on Monday to block OpenAI from training new models without independent oversight and to cut minors off from ChatGPT.

The motion is an emergency request for an injunction. It asks the court to impose three categories of restraint on the company: halt new model development unless independent safety guardrails are in place, stop collecting children's data, and end marketing claims and product behavior that Uthmeier's office considers deceptive. The filing escalates a legal fight that began this past June, when Uthmeier's office sued OpenAI. That lawsuit followed a criminal investigation into the company launched in April, in the aftermath of the 2025 mass shooting at Florida State University. The suspect in that case allegedly used ChatGPT in the lead up to the shooting.

Uthmeier announced the motion in a video shared on X, laying out his demands in blunt terms.

"No new model development without independent safety guardrails; no more harvesting children's data; no more calling this product safe, accurate, or reliable; no more pretending it's human; no more tricks designed to keep users talking past the point of danger," Uthmeier said Monday. "If Sam Altman meant what he said about slowing down, he can join our ask to the court. If he will not, we ask the court to do what OpenAI will not do for itself — protect Florida families."

OpenAI did not immediately respond to a request for comment on the new motion.

The company has previously addressed the underlying investigation. "Last year's mass shooting at Florida State University was a tragedy, but ChatGPT is not responsible for this terrible crime... ChatGPT is a general-purpose tool used by hundreds of millions of people every day for legitimate purposes. We work continuously to strengthen our safeguards to detect harmful intent, limit misuse, and respond appropriately when safety risks arise," OpenAI said in a statement after Uthmeier's office first opened the criminal investigation in April.

The injunction built on OpenAI's own words

The most unusual feature of Monday's filing is its argumentative structure. Uthmeier's office grounds the injunction request in OpenAI's own recent public statements and disclosures — not only in the allegations about the Florida State case.

The brief cites OpenAI's recent security disclosures, including one made after the hacking of Hugging Face, as evidence of unacceptable risk. It also points to Sam Altman's own calls for the AI industry to slow the pace of frontier model development as support for court intervention.

"It is a rare request for an injunction where the Defendants themselves have publicly endorsed it," Uthmeier writes in the brief. "They have asked the government to tie them to the mast. Plaintiff brings good news to the Defendants: The Florida Attorney General is answering your cry for help with a motion to enjoin you from harming Floridians with your reckless, unacceptably risky product."

That framing gives the case significance beyond Florida. If a state court accepts the argument that a company's public calls for regulation can justify judicial restraints on its product development, it would set a precedent for how state attorneys general use corporate safety statements against AI developers. It would also test whether courts are willing to impose development-stage limits on model training — a remedy that goes well beyond the content moderation and age-verification demands that have dominated state-level AI litigation so far.

The request to bar minors from ChatGPT entirely adds a second front. Age-gating and data collection practices for AI chatbots have drawn scrutiny from multiple regulators, and a court order on this point would directly affect how OpenAI operates one of the most widely used consumer AI products in the United States.

A pattern of scrutiny after mass shootings

Florida's lawsuit is not the first time OpenAI has been connected to a mass shooting. In February, Canadian regulators called for OpenAI to change how it approaches threats of harm. That intervention followed a Wall Street Journal report claiming the company flagged the account of alleged Tumbler Ridge shooter Jesse Van Rootselaar in 2025 but failed to bring the threats to law enforcement. Following that tragedy, OpenAI agreed to new policies around how it works with Canadian law enforcement.

The Canadian episode established a pattern: a violent incident involving an alleged ChatGPT user, followed by regulatory pressure, followed by policy changes at the company. Florida's motion pushes the pattern further. Rather than negotiating new reporting or escalation procedures, Uthmeier is asking a court to freeze the company's core activity — training new models — until independent oversight exists.

The distinction matters for the broader industry. OpenAI's commitments in Canada covered how the company handles threats of harm once detected. The Florida injunction, if granted, would reach the development process itself, before any product reaches users. For competitors racing to release frontier models, a Florida court order conditioning model training on external safety review would become an immediate reference point in their own regulatory and legal exposure.

The FSU shooting gives the case its emotional and political weight. The suspect allegedly used ChatGPT in the lead up to the attack, and Uthmeier has built both a criminal investigation and now a civil injunction request on that connection. OpenAI's defense rests on the product's scale and purpose: a general-purpose tool used daily by hundreds of millions of people, with safeguards the company says it strengthens continuously.

What happens next depends on the state court. Uthmeier's brief dares OpenAI to endorse the injunction voluntarily, framing Altman's slowdown rhetoric as a public endorsement the company cannot retract. If OpenAI opposes the motion — as its prior statements suggesting it will — the company will have to argue against restraints it has itself described as necessary for the industry. Either way, the Florida case now poses the sharpest test yet of whether public calls for AI regulation can be converted into binding legal limits on frontier model development.