OpenAI's Legal Pressure Escalates as State AGs Investigate and Florida Sues Altman

OpenAI is suddenly fighting on a new front: state law. A coalition of US state attorneys general has opened an investigation into the company and served it with a subpoena on Friday, days after Florida became the first state to sue OpenAI and CEO Sam Altman personally. The accusations are serious, the timing is awkward, and Altman's own name is now on a lawsuit.

Florida fired the first shot. Attorney General James Uthmeier filed an 83-page complaint accusing OpenAI of deceptive practices, of marketing ChatGPT aggressively, including to children, while concealing risks and suppressing internal safety warnings. The complaint goes further than corporate liability, it seeks to hold Altman personally responsible, citing what it calls an utter disregard for the risk to human life. Some of the claims are dramatic, including that ChatGPT contributed to deadly rampages and pushed some users toward suicide.

Then it widened. A group of state attorneys general has now opened a joint investigation, and OpenAI was served Friday with a subpoena demanding documents across a broad range: advertising, how it drives user engagement and retention, its handling of consumer and health data, its treatment of minors and seniors, and even model sycophancy, the tendency of a model to tell users what they want to hear. OpenAI said on Friday it is engaging constructively with the AGs.

This is a different kind of risk than the usual AI debate. Most AI policy fights are about national security or copyright. This one is consumer protection, run by elected state officials, aimed at how the product affects ordinary people and kids. And the personal-liability angle is unusual, going after the CEO by name raises the stakes for every AI founder watching. It also lands while OpenAI is preparing to go public, exactly when a clean legal story matters most.

Worth keeping perspective. These are allegations, not findings, and the most extreme claims, that a chatbot helped cause mass violence, will be hard to prove and fiercely contested. OpenAI says it takes safety seriously and is cooperating. But even an investigation creates real overhang: subpoenas, discovery, headlines, and a coalition of AGs is hard to settle quietly. The legal cost may end up smaller than the reputational one heading into an IPO.

So the AI legal map is filling in fast. The federal fight is about Anthropic's models, and now the state fight is about OpenAI's users. Two of the biggest labs, two very different governments, both leaning on the same lesson: the bigger AI gets, the more the law shows up. Watch whether more states join Florida, because that is what turns a lawsuit into a problem.