Florida sues OpenAI, Altman over ChatGPT risks

Florida sued OpenAI and CEO Sam Altman on Monday, moving the dispute over chatbot safety from policy hearings into a state consumer-protection case that investors may have to treat as a legal-risk event.

In the complaint, Florida Attorney General James Uthmeier accuses OpenAI and Altman of deceptive practices and unsafe marketing around ChatGPT. His office described it as the first state-led lawsuit against the company. At the centre is an allegation that OpenAI presented ChatGPT as broadly safe for consumers, including minors, while withholding or softening warnings about serious risks.

Investors have a reason to watch the legal theory, not just the venue. Florida is trying to tie liability to the way a frontier-model company describes, governs and sells a general-purpose AI product. If that argument survives, the case would sit near privacy enforcement, social-media child-safety litigation and product-liability claims, even with the facts still untested.

“Today, we announced the first-in-the-nation state-led lawsuit against OpenAI and its CEO, Sam Altman.”

James Uthmeier, Florida attorney general

OpenAI, in a statement cited by AP News, rejected the premise. It said ChatGPT is a general-purpose tool used for legitimate purposes by hundreds of millions of people. The company has said it keeps improving safeguards and responding to misuse.

“ChatGPT is a general-purpose tool used by hundreds of millions of people every day for legitimate purposes.”

OpenAI statement

According to Reuters, the complaint cites alleged harms involving minors and refers to two shootings in its account of the risks. AP reported that one case referenced in the wider dispute involved Adam Raine, who was 16. Such details give the filing political force, but they also narrow the legal question: can downstream harm from chatbot interactions be tied to the company’s public claims, design choices or executive conduct?

A state case with market stakes

For private AI groups, the lawsuit lands at an awkward point. OpenAI is not public, yet its valuation, partnerships and eventual listing prospects depend partly on whether investors can model its regulatory exposure. A state complaint naming Altman personally makes that exercise less straightforward before any court reaches liability.

Public-market investors usually price uncertain legal regimes with a discount. Should Florida’s theory become a template for other states, the risk is no longer one damages claim in one court. It becomes a possible patchwork of actions over product disclosures, safety testing, youth protections and governance promises, with compliance costs rising as the growth story gets harder to present cleanly to strategic backers and private-market investors.

Uthmeier’s office also chose a relatively tight frame. It cast the case as consumer protection, not a broad referendum on AI. That may help Florida argue it is challenging representations and conduct rather than trying to regulate speech or software in the abstract.

What OpenAI has to defend

A likely first defence is that ChatGPT is used across education, business, programming and research, rather than designed for the narrow harms Florida cites. OpenAI may also point to safety work, warnings and model updates as evidence of mitigation rather than deception.

Disclosure is the harder issue. Florida’s complaint turns on what users were told, what risks the company allegedly understood, and whether Altman’s public role makes him personally accountable. For a company built around rapid adoption, the case tests the space between product ambition and risk language. Courts have treated that gap seriously in other consumer-technology disputes when marketing claims ran ahead of internal controls.

Florida still has to prove its case. OpenAI will have procedural and factual arguments, and courts are only starting to decide how older consumer-protection statutes apply to generative AI. The filing still offers regulators a map: focus on safety representations, link them to vulnerable users and point to senior leaders when public assurances are central to a company’s pitch.

The near-term market read is less about damages than optionality. One state case can be settled, narrowed or dismissed. A repeatable legal theory can travel. If other attorneys general follow Florida, OpenAI and its rivals may have to show that safety governance is auditable, not merely improving.