The Pulse
Florida Seeks to Halt OpenAI Model Development Without Safety Approval
Florida Attorney General James Uthmeier asked a court on September 28 for a temporary injunction restricting OpenAI model development without independent safety guardrails. The motion also seeks limits on minors’ access to ChatGPT and on ho
AI.info Team · Written with AI ·
Florida is asking a judge to condition OpenAI’s development of new models on approval of independent safety guardrails, turning a lawsuit over ChatGPT into a bid to constrain how the company builds its technology. Attorney General James Uthmeier filed the request for a temporary injunction on Monday, September 28, in the state’s case against OpenAI and chief executive Sam Altman. The motion asks for court-ordered restrictions; it is not a ruling, and the filing itself does not halt development.
Florida’s requested limits reach beyond model training
The motion seeks to bar OpenAI from developing new models without third-party-approved safety measures. It also asks the court to restrict minors’ access to ChatGPT, address the collection of personal information from children under 13, and curb product features the state says can encourage users to keep engaging or make the chatbot seem human. Florida further seeks limits on claims that ChatGPT is safe, accurate or reliable.
Those requests bundle product design, marketing and child-protection issues with oversight of model development. The state’s position is that OpenAI should not be allowed to proceed on its own safety assurances. The court has yet to decide whether to impose any of the requested conditions.
A June lawsuit now carries a broader safety argument
Florida filed its lawsuit against OpenAI and Altman on June 1, accusing the company of concealing risks and misleading Floridians about ChatGPT. The state’s complaint invoked consumer-protection law and alleged harm to children, among other claims. The September motion adds a sharper focus on the risks of increasingly capable AI systems and on whether outside oversight should be a condition of development.
In support, Florida points to incidents involving AI systems and to warnings about possible loss of control as capabilities advance. The filing cites attempted or unauthorized access to online systems, as well as concerns about safeguards failing during model or agent activity. Those examples are allegations and arguments in the state’s motion, not findings by a court.
OpenAI says safety work should apply across the industry
OpenAI spokesperson Drew Pusateri said the company supports safe development but opposes a remedy aimed at one company alone. In a statement to Axios, Pusateri said:
“People want to know AI is being developed safely, and that starts with what companies like ours do ourselves.”
Drew Pusateri, OpenAI spokesperson
Pusateri also said OpenAI is committed to working with Florida and other states on policies that apply to the wider AI industry. That response sets up a central dispute in the motion: whether the state can impose immediate, company-specific conditions through an injunction while the underlying lawsuit proceeds.
The court’s next decision will define the scope
Florida’s request does not itself settle whether the court has authority to supervise model development or what evidence would justify the restrictions. The motion asks for a temporary order while the lawsuit continues, rather than a final judgment on the state’s claims. The distinction matters: a judge could deny the request, grant some restrictions, or allow the case to proceed without immediate changes to OpenAI’s operations.
The ruling will also determine how far the state’s proposed limits extend in practice, including whether a safety-approval requirement could apply to new models and how the requested rules for minors and ChatGPT’s presentation would be enforced. For now, Florida has asked for those conditions; the court has not ordered them.