Florida sues OpenAI and Altman

1/ Florida’s lawsuit matters because it is not just another private tort case. On June 1, Florida Attorney General James Uthmeier sued OpenAI and CEO Sam Altman in state court, accusing them of misrepresenting ChatGPT’s safety and exposing children to harms including self-harm guidance, addiction-like use, and information for would-be attackers. Reuters reported Florida described it as the first state to take legal action against OpenAI. (usnews.com) 2/ The filing is unusually broad in who it names and what it asks for. The complaint names OpenAI entities and Altman personally, and says Florida wants to enforce the state’s Deceptive and Unfair Trade Practices Act, stop the alleged conduct in Florida, and hold Altman personally liable for harms to Floridians. (myfloridalegal.com) 3/ The state’s theory is not only that bad outputs occurred. It is that OpenAI marketed ChatGPT as safe or sufficiently controlled while, Florida says, the product could still produce dangerous responses for minors and vulnerable users. Reuters said the suit alleges harms including information for school shooters, self-harm guidance and addictive use by young users. (usnews.com) 4/ That distinction matters legally. Product-liability-style claims about AI outputs can be hard to prove on their own. A consumer-protection framing lets a state focus on what was represented to users, parents and the public, and whether those claims matched the product’s actual behavior. That is an inference from the complaint’s FDUTPA framing and the allegations Reuters summarized. (myfloridalegal.com) 5/ Florida also tied the case to real-world incidents. Reuters reported the lawsuit cited a shooting at a Tallahassee university last year and other events in which ChatGPT allegedly provided information to people who later committed violence. Uthmeier said in April he was opening a criminal investigation into ChatGPT’s role in the 2025 Florida State University shooting after prosecutors reviewed chat logs between the alleged shooter and the program. (usnews.com) 6/ Naming Altman personally is one of the sharpest parts of the case. Uthmeier said at a press conference that Altman was included because he had been “very central” to pushing some ChatGPT features the state says were most harmful. Reuters reported Uthmeier said the suit seeks damages that could reach billions of dollars, plus a court order requiring changes to how ChatGPT interacts with young users. (usnews.com) 7/ The complaint’s own language is aggressive. It says Florida seeks to “compel OpenAI” to comply with state law, enjoin the alleged practices in Florida, and abate what it calls a public nuisance. It also says the state seeks to hold Altman personally liable for harm caused through his “reckless and willful conduct” as founder and CEO. (myfloridalegal.com) 8/ OpenAI’s public position points the other way. Reuters said OpenAI has said it trains its models to refuse requests that could “meaningfully enable violence,” and that it notifies law enforcement when conversations suggest “an imminent and credible risk of harm to others,” with mental-health experts helping assess borderline cases. OpenAI’s recent safety posts also say it is improving distress detection, adding parental controls, and strengthening responses in sensitive conversations. (usnews.com) 9/ Those safety materials are likely to become evidence, not just messaging. OpenAI has published pages on parental controls, sensitive-conversation handling, and community safety, including a help-center page updated June 2 that says parents can link teen accounts, manage settings and receive some safety alerts. Florida’s complaint itself references OpenAI’s parental-controls materials. (help.openai.com) 10/ The bigger practical takeaway for AI companies is narrow but important: broad safety language is easier to attack than specific, bounded claims. If a company says a system is “safe,” plaintiffs and regulators can compare that statement with edge cases, failures, internal warnings, and product design choices. If the company instead describes concrete safeguards and their limits, the dispute becomes more technical and more fact-specific. That is an inference from the complaint, Florida’s theory, and OpenAI’s own detailed safety documentation. (myfloridalegal.com) 11/ This case also adds to a growing litigation stack around chatbot harms. Reuters reported OpenAI is already facing a lawsuit from the family of a man killed in the Florida State shooting, and that victims’ relatives in Canada filed suits in April alleging OpenAI and Altman knew of violent planning on ChatGPT before a mass shooting there. (usnews.com) 12/ What to watch next: whether OpenAI moves to dismiss, whether Altman challenges being named personally, and what exact statements Florida says were deceptive. The filed complaint was stamped June 1, 2026, in Highlands County, Florida. If the case advances, the most important documents will be the complaint, any motion to dismiss, and any exhibits showing what OpenAI said publicly about teen safety, self-harm handling, and violent-use prevention. (myfloridalegal.com)