The State of Florida filed an 83-page civil lawsuit against OpenAI and its chief executive Sam Altman in the 10th Judicial Circuit this week, making it the first state in the country to take the ChatGPT maker to court over child safety claims. The filing does not read like a routine regulatory action. It reads like a prosecution — one that alleges a company built something it knew was dangerous, shipped it anyway, and spent years telling the public the opposite.

At the core of the complaint is a charge that should alarm anyone who has handed a teenager a smartphone: that OpenAI designed ChatGPT with addictive mechanics, concealed the psychological risks from parents and regulators, and allowed the product to reach minors without meaningful safeguards. The state's lawyers use the word "seduction" — not metaphorically, but as a description of how the system is alleged to have been engineered to keep users engaged regardless of the harm that engagement caused.

Two deaths anchor the most damning sections of the lawsuit. Florida's complaint points to a pair of fatal shootings and draws a direct line to ChatGPT interactions preceding those incidents. The state stops short of claiming the chatbot pulled a trigger, but it does allege that the product's outputs — unchecked, unfiltered, and actively engaging vulnerable users — were a material factor in both tragedies. The specifics of those cases, laid out in the filing, will be the legal and factual terrain on which this lawsuit is won or lost.

The state's theory of liability rests on several overlapping claims: that OpenAI violated Florida consumer protection law by misrepresenting the safety of its product; that the company failed in its duty to warn users — particularly minors and their parents — of known psychological risks; and that Altman himself bears personal responsibility for decisions made at the executive level about how and when to disclose those risks. Naming a CEO personally in a state consumer protection action is an aggressive move, and it signals that Florida's attorneys general office intends to make this costly and uncomfortable for the company's leadership, not just its legal department.

The addictive-design allegation is worth sitting with. It mirrors the legal theory that proved effective in litigation against social media platforms — the argument that engagement-maximizing design is not a neutral technical choice but a deliberate product decision with foreseeable human costs. Florida is betting that the same logic applies to large language models: that a system trained to be fluent, responsive, emotionally resonant, and always available is not accidentally compelling to a lonely or troubled teenager. It is, the state argues, compellingly dangerous by design.

OpenAI has not yet filed a formal legal response to the complaint, and the company's public posture has been to emphasize its safety investments and the controls it claims are built into the product. That defense will face scrutiny, because the lawsuit specifically addresses the gap between what the company said publicly about safety and what it allegedly understood internally. That gap — between the press release and the internal assessment — is where these cases tend to be decided.

Sam Altman's inclusion in the complaint is pointed. He has spent years positioning himself as a thoughtful steward of powerful technology, testifying before Congress about responsible AI development and volunteering his company for regulatory oversight while simultaneously pushing the product's capabilities as fast as the hardware will allow. Florida is, in effect, arguing that the responsible-steward narrative was a performance — and that the actual decisions made by the company's leadership prioritized growth and revenue over the safety of the children using the product.

This lawsuit will not resolve quickly. Florida's legal action opens a discovery process that could force OpenAI to produce internal communications, safety assessments, product design documents, and executive correspondence that the company has never been required to make public. That is, in many respects, the strategic value of the filing regardless of how it ultimately resolves at trial. The documents that come out of litigation like this have a way of becoming the story.

Other states are watching. The precedent Florida is trying to set — that a generative AI company can be held civilly liable under state consumer protection law for harms caused by its product to minors — is one that dozens of state attorneys general have been evaluating since ChatGPT became a household name. If Florida's legal theory survives early motions to dismiss, expect the map of OpenAI defendants to expand considerably before the end of the year.