Florida Takes OpenAI to Court Over ChatGPT Safety Claims

Florida's attorney general has filed what he calls the first state-level lawsuit against OpenAI and CEO Sam Altman. The complaint alleges that the company released ChatGPT to the public while hiding serious safety risks, and that it downplayed warnings from its own experts. The 83-page lawsuit, filed in Highlands County, seeks damages for Florida residents and wants to stop what the state views as deceptive marketing practices.

This legal action marks a significant step in how states are responding to generative AI — the technology behind ChatGPT that can write text, answer questions, and create other content. Florida's attorney general had already opened a criminal investigation in April before deciding to pursue this civil (non-criminal) lawsuit.

What Florida Is Accusing OpenAI Of Doing

The lawsuit centers on a straightforward claim: that OpenAI knew ChatGPT had potential dangers but marketed it aggressively anyway without telling the public. Specifically, Florida alleges that the company ignored internal warnings about risks — including the possibility that ChatGPT could be used to encourage self-harm or violence — and that it hid these concerns from users, including young people.

The complaint also points to research from Common Sense Media showing that one in three teenagers use AI chatbots to talk about relationships and personal issues. This detail matters because the lawsuit frames it as evidence that vulnerable users are relying on ChatGPT without understanding its limitations.

The core legal argument is that OpenAI prioritized getting to market quickly and generating revenue over protecting users' safety.

The Lawsuit Comes With Other Legal Troubles

OpenAI already faces individual lawsuits. Parents of people killed in a shooting at Florida State University have sued the company, claiming ChatGPT was involved. In California, the family of an 83-year-old woman sued after her son, who had become increasingly paranoid and delusional while using ChatGPT, killed her. They argue the chatbot made his condition worse.

Sam Altman, OpenAI's CEO, is also defending himself in a separate lawsuit filed by Elon Musk, who wants Altman removed from leadership and accuses him of abandoning their original vision for the company.

The timing matters. States across the country are starting to treat AI as a major regulatory issue, particularly where federal rules either don't exist yet or are still being written.

Why This Legal Strategy Could Matter

The Florida lawsuit uses an approach we have seen work in the past — looking for internal company documents and communications that show the company knew about problems but hid them anyway. This worked in lawsuits against social media companies, where internal research on user harms became key evidence. The pattern is recognizable to anyone who followed litigation around Facebook and TikTok over the past decade.

Florida's lawyers are using state consumer protection laws, which give state officials broad power to sue on behalf of their residents. This is often faster and more flexible than waiting for federal regulators to create new AI rules. The complaint's focus on marketing claims and concealed risks fits neatly into tools states already have.

The Florida action establishes something that other state attorneys general will likely notice: a detailed roadmap for suing AI companies. Other states could adapt these arguments for lawsuits of their own, particularly around disclosure of safety concerns.

From a technical standpoint, the lawsuit raises a question that the AI industry has largely avoided: what does a company actually owe users when it knows an AI system has limitations or potential harms. Right now, most AI companies rely on terms of service and use policies — basically, the fine print — to limit their liability. State enforcement could change that, requiring companies to actively disclose what they know about risks rather than burying the information in legal documents.

What This Signals for the Broader AI Industry

There is a pattern emerging here worth taking seriously. The AI industry has publicly embraced the idea of "AI safety" as important, but the Florida lawsuit suggests that courts and regulators will judge companies by their actions — by what they knew internally and what they chose to tell (or not tell) the public.

Some former OpenAI employees and AI safety researchers have already raised concerns that pressure to move quickly and compete for users can undermine safety work inside these companies. The Florida lawsuit, if it proceeds, will test whether those concerns have legal weight.

I spent the 1990s covering the early internet, and I have seen this tension before. Web browsers arrived fast and rough. Social media platforms grew explosively before anyone really understood the harms. The pattern repeats: companies move quickly to capture users and market share, and later we learn about risks they either didn't fully explore or didn't fully disclose. The difference with AI may be the sheer scale and speed of adoption, and the fact that we are still learning what these systems can and cannot do safely.

The Florida lawsuit will almost certainly prompt AI companies to review their internal safety documentation and think harder about what they tell users. Other states are watching. The outcome could change what "responsible AI deployment" means in practice — not just as marketing language, but as something with legal teeth.

This is one of those cases where the legal and technical questions are running ahead of settled answers. What happens in court will likely shape how AI companies operate, probably before federal regulators have finished writing rules.