Eighty-three pages, ten counts, and a CEO named personally. On June 1, 2026, Florida became the first U.S. state to haul OpenAI and Sam Altman into court, and the target isn't the technology itself but how it was sold. The complaint opens a new front: product liability applied to artificial intelligence, arriving precisely as major AI labs race toward public markets, as seen with the Anthropic IPO filing.
What Florida Is Actually Alleging
Filed by Attorney General James Uthmeier at the Highlands County Circuit Court, the complaint lists ten counts. Four involve deceptive and unfair trade practices under the Florida Deceptive and Unfair Trade Practices Act. Two allege negligence, two product liability, one fraud, and one public nuisance. The core argument is that OpenAI released and aggressively marketed ChatGPT to the public, including minors, while concealing serious risks and suppressing internal safety warnings. Penalties quantified by the AG's office run into the billions, and the complaint demands an order blocking data collection from minors without genuine parental consent.
Why Altman Is Named Personally
Functionally, the heaviest legal weight sits on personal liability. The action targets Altman directly, for his conduct as founder and CEO, using unambiguous language: the complaint describes “total disregard for the risk to human life.” Naming the head of a tech company of this scale personally is rare. If the theory holds, AI lab executives won't be able to shelter behind the corporate veil when a product causes harm. That precedent alone matters more than any dollar figure attached to a penalty.
Twenty-Plus Suits and a Criminal Probe
Florida's lawsuit didn't emerge from nowhere. According to court records, more than twenty private actions have already been filed against OpenAI for harms linked to ChatGPT use, including suits brought by families of shooting victims and by relatives of people who died by suicide or developed delusions after using the chatbot. Running in parallel, Uthmeier opened a separate criminal investigation in late April 2026, still ongoing, distinct from this civil case. OpenAI has responded publicly that minors need strong protections and that the company has put safeguards and industry policies in place.
What Changes for AI Product Liability
This is where the story becomes relevant for anyone building technology. AI companies have consistently argued that a model is a neutral tool. Treating it as a potentially defective product, under product liability rules, flips the burden of proof. Europe is taking a different route: preventive regulation by risk level under the AI Act, implemented nationally as Spain did with its organic law and as already weighing on AI agents for freelancers and SMEs. Florida is trying the courtroom, case by case. Two philosophies, one target.
The 10 Counts in Florida’s Complaint
Source: civil complaint, Highlands County Circuit Court, June 1, 2026
Uthmeier has stated publicly that he expects other states to follow. If they do, June 1, 2026 won't remain a footnote about the Sunshine State. It will mark the date when AI product liability stopped being an academic hypothesis and became a concrete legal risk, with a specific name at the top of the filing.
