Florida Sues OpenAI. The Tobacco Playbook Is Back.

Florida filed the first state lawsuit against OpenAI on June 1, 2026. The legal theory it chose is not new. It is the same one that dismantled the tobacco industry.

Florida Sues OpenAI. The Tobacco Playbook Is Back.

Product liability law was built for physical things. Cars with faulty brakes. Pharmaceutical drugs whose clinical trial data companies controlled and withheld. Cigarettes whose manufacturers documented the harms, buried the research, and kept selling. The legal framework that eventually worked against Big Tobacco was built on a single claim: the danger was known before the product reached the market. On June 1, 2026, Florida applied that claim to a chatbot.

Attorney General James Uthmeier filed an 83-page civil complaint against OpenAI and CEO Sam Altman in Highlands County Circuit Court, becoming the first US state to bring a civil action against the company. The complaint names five OpenAI corporate entities and Altman personally, arguing he suppressed internal safety warnings in pursuit of market dominance. Florida's attorney general has said OpenAI could face liability running to potentially billions of dollars if found responsible.

What the complaint actually says

The accusations are wide-ranging. Florida alleges ChatGPT assisted a gunman who killed two people at Florida State University in April 2025, after prosecutors reviewed the shooter's chat logs with the tool. The complaint also references a February 2026 mass shooting in Tumbler Ridge, Canada, where families of seven victims are pursuing separate civil claims against OpenAI.

Florida further alleges the product addicts minors to a tool with no age verification, encourages vulnerable users toward self-harm, and collects children's data without parental consent in breach of the FTC's COPPA rules. According to a Common Sense Media study cited in Florida's complaint, 72% of teenagers have used AI for companionship, and one in three use chatbots for deep social interaction.

OpenAI's response is familiar. A spokesperson acknowledged AI is a powerful technology requiring guardrails for minors, pointing to existing safety features. The company had previously denied responsibility for the Florida State shooting. Neither response addresses the structural argument Florida is making.

The design defect framing

Strict product liability for design defects is the detail that separates this lawsuit from everything that came before it. Private family lawsuits against AI companies have repeatedly struggled against Section 230 of the Communications Decency Act, which broadly shields online platforms from liability for third-party content. Florida's design defect count sidesteps Section 230 entirely: the state is arguing the product itself is defective by design, and that OpenAI knew this before it went to market.

That is exactly how states eventually pierced the tobacco industry's legal defences. The product liability cases that mattered were not about individuals choosing to smoke. They were about companies that conducted internal research, documented the harms, and continued selling. Florida's complaint mirrors this structure explicitly: ChatGPT's valuation grew from $17 billion to over $850 billion in under four years, the state argues, built on concealed safety data and aggressive marketing to users the company knew were at risk.

Italy's competition authority reached a parallel conclusion through a different route, requiring AI companies to warn users about hallucinations under existing consumer protection law, without waiting for AI-specific legislation to catch up.

Why timing matters

The lawsuit lands weeks before OpenAI's anticipated IPO at a reported valuation between $852 billion and $1 trillion. A state attorney general pursuing personal liability against a CEO in a first-of-its-kind action is a structurally different legal risk from the family lawsuits that preceded it. The design defect theory's survivability against a Section 230 challenge remains genuinely unresolved — it has not been tested at this level before. If it holds, every AI company with a consumer product faces the same exposure.

Florida has not proven its case. The lawsuit is 83 pages of allegations, not findings. But the legal architecture it introduces does not require a conviction to do damage. The tobacco analogy holds here too: the legal theory reshaped corporate behaviour long before the settlements were signed.


Editor's note

Every piece published on The Bright Minded goes through careful verification, but mistakes can happen. If you spot an error, have additional information, or want to flag anything, write to rosalia@thebrightminded.com.