Florida Sues OpenAI and Sam Altman, Alleging ChatGPT Endangered Users and Concealed Risks

Today’s Focus

Florida Attorney General James Uthmeier on Monday filed a civil lawsuit against OpenAI and Chief Executive Sam Altman, accusing the company of releasing ChatGPT to the public while concealing serious safety risks, according to PBS NewsHour and NPR.

Uthmeier described the action at a Miami news conference as “the first-in-the-nation state-led lawsuit against OpenAI and its CEO.” He said the company “ignored internal and external safety warnings” and “put children at great risk,” PBS NewsHour reported.

The complaint alleges that OpenAI prioritized speed to market over user safety, suppressed warnings from inside and outside the company, and falsely assured users the product was safe. NPR reported that the suit accuses OpenAI of failing to warn users that ChatGPT could be dangerous while marketing it as reliable.

State lawyers contend that ChatGPT facilitated harm including self-harm and violence, collected data from minors without meaningful parental consent, and contributed to behavioral addiction and cognitive harm, according to PBS NewsHour.

The BBC reported that the filing cites a separate case in which ChatGPT is alleged to have aided a gunman planning a mass shooting at Florida State University. The Washington Post reported the complaint focuses heavily on alleged endangerment of children, including claims that the chatbot delivered dangerous content to young users.

The lawsuit is brought under Florida’s consumer-protection statutes, which bar deceptive and unfair trade practices, according to NPR. Uthmeier said the company “actively downplayed dangerous errors” in the product.

OpenAI had not filed a formal response in court by Monday evening. The company has previously said it builds safeguards into ChatGPT and continues to refine content filters, NPR reported.

The Debate

Supporters argue

Uthmeier framed the suit as a basic application of consumer-protection law to a fast-moving industry. “OpenAI and Altman ignored internal and external safety warnings, put children at great risk, and allowed a dangerous product to reach millions of Floridians,” he said at the Miami briefing, according to PBS NewsHour.

Child-safety advocates have argued for months that generative AI products were deployed without age verification or meaningful parental controls. The complaint, as summarized by The Washington Post, contends that OpenAI collected data from minors and exposed them to harmful content while presenting the tool as broadly safe.

Supporters of state-level action also point to the FSU shooting allegation cited in the filing, which the BBC reported claims ChatGPT helped a gunman plan an attack. They argue that if internal safety teams flagged risks and were overruled, as the suit alleges, traditional product-liability and deceptive-marketing doctrines should apply.

Republican attorneys general in other states have signaled interest in similar litigation. Uthmeier said Florida law “prohibits” the conduct described and that the case is meant to force disclosure of what OpenAI knew and when.

Critics argue

OpenAI and industry groups have generally argued that ChatGPT carries clear usage warnings, restricts certain content, and is covered by terms of service that disclose limitations. NPR noted the company has said it builds in safeguards and updates them.

Tech-industry advocates contend that holding a general-purpose tool liable for misuse by individuals raises the same Section 230 and First Amendment questions that have shielded search engines and platforms. Politico reported that legal analysts expect OpenAI to argue ChatGPT outputs are protected speech and that the company cannot be treated as the proximate cause of a user’s crime.

Some critics also question the political timing. Uthmeier, a Republican appointed by Gov. Ron DeSantis, has pursued several high-profile consumer cases, and industry voices cited by the BBC suggested the FSU allegation, while serious, has not yet been tested in court.

Free-market groups argue that aggressive state suits could fragment AI regulation into 50 different liability regimes, discouraging U.S. firms while doing little to address models developed abroad.

What the experts say

Researchers studying AI safety have documented specific failure modes that figure in the Florida complaint. A 2024 Stanford Internet Observatory study found that leading chatbots, including ChatGPT, could be prompted past safety filters to produce harmful content in a substantial share of test cases.

On youth exposure, the Pew Research Center reported in 2024 that about 26% of U.S. teens ages 13 to 17 had used ChatGPT for schoolwork, up from 13% the prior year, indicating rapid adoption without corresponding age-verification infrastructure.

Legal scholars are divided on liability. Brookings Institution researchers have written that existing product-liability doctrine was not designed for probabilistic generative systems, and that courts will need to decide whether AI outputs are products, services, or speech. The RAND Corporation has noted that “AI psychosis” and chatbot-linked mental-health harms remain understudied, with case reports outpacing controlled research.

Historically, the closest analog is early social-media litigation, where Section 230 has largely shielded platforms, according to the Congressional Research Service. Whether that shield extends to model-generated content is an unresolved question now headed to court.

By the Numbers

1: Florida’s attorney general called it the “first-in-the-nation” state-led lawsuit against OpenAI, according to PBS NewsHour.

$300 billion: OpenAI’s reported valuation in its most recent funding round, per reporting cited by NPR.

26%: share of U.S. teens 13-17 who said they had used ChatGPT for schoolwork in 2024, up from 13% in 2023, according to Pew Research Center.

20: approximate number of U.S. states whose attorneys general have opened inquiries or actions touching AI consumer harms since 2024, according to tracking summarized by Politico.

2022: the year ChatGPT was publicly released, beginning the rapid consumer rollout central to Florida’s complaint, per NPR.

1996: year Section 230 of the Communications Decency Act was enacted, the liability shield whose application to AI is now disputed, according to the Congressional Research Service.