Florida sues OpenAI and Sam Altman over alleged user exploitation and safety failures

Florida Attorney General James Uthmeier has filed a civil lawsuit against OpenAI and its chief executive, Sam Altman, alleging that the company engaged in deceptive trade practices, negligence, and product liability violations. The complaint accuses the artificial intelligence firm of exploiting users, including residents of Florida, to boost its market value while disregarding significant safety risks. The suit seeks financial penalties and court orders, explicitly attempting to hold Altman personally liable for what it describes as reckless and willful conduct.

The legal action centres on allegations that OpenAI prioritised engagement and data collection over user safety. The complaint argues that the company’s systems exhibit a propensity for sycophancy, agreeing with users to increase interaction and training data. Uthmeier stated that the lawsuit aims to hold Altman accountable for harm caused to Floridians through an "utter disregard for the risk to human life." This civil suit is distinct from a separate criminal investigation into the company that the Attorney General opened earlier this year, which remains ongoing.

The filing cites specific violent incidents where individuals allegedly used ChatGPT during the planning stages of attacks. These include a mass shooting at Florida State University last year, where the alleged shooter reportedly consulted the chatbot for advice on weapons and media attention. Additionally, the suit references the shooting of two University of South Florida students earlier this year, where the alleged shooter allegedly received guidance on hiding bodies. In both cases, OpenAI has denied responsibility, stating that the chatbot provided factual responses available in public sources.

Beyond Florida, the complaint highlights a mass shooting in British Columbia in February, where eight people were killed. The alleged shooter was in regular contact with ChatGPT, which flagged the account for violence but only deactivated it rather than alerting authorities. The shooter subsequently created a second profile to continue the conversation. The suit also notes that OpenAI faces at least eight lawsuits stemming from incidents of mass violence or self-harm, including cases involving suicide planning.

The legal document further criticises OpenAI’s advertising for failing to disclose that the software can provide false, nonsensical, or hallucinated information. Uthmeier argues that the company’s systems present a danger of addiction, cognitive decline, and violence. OpenAI has previously stated that it designs its systems with safety at every step and employs safeguards to guide users toward real-world support during sensitive conversations, though the Attorney General contends that real-world events suggest otherwise.