Florida has become the first state to sue OpenAI and Sam Altman over alleged harms from ChatGPT.
The lawsuit, filed by Attorney General James Uthmeier, accuses OpenAI of rushing unsafe products into the world, marketing ChatGPT as safe and reliable, and ignoring warnings about how the chatbot could affect children, vulnerable users, and people planning violence. The complaint cites alleged links between ChatGPT and multiple violent incidents, suicides, and self-harm cases.
The question for the court is whether OpenAI can be held liable when users lean on ChatGPT for dangerous, delusional, or violent plans. The company will argue that its models are general-purpose tools with safeguards. Florida is arguing that the tool was built and sold in a way that made predictable harms inevitable.
Chatbots are intimate, persuasive, always available, and very good at sounding confident even when they’re completely wrong. That combination is bad enough when the output is fake case law. It is much worse when the user is a teenager, a person in crisis, or someone already halfway into a violent fantasy.
The lawsuit may overreach. It may also be the start of a much bigger fight over whether AI companies get to call these systems companions, tutors, therapists, coders, search engines, and productivity tools, then shrug when people use them as intended.
Florida government probably not the best poster child for this, but someone with resources likely should go after "move fast, break people's brains."
I've had the impression, the creators of LLM "AI" were it's first victims. The concept of "AI as a cult" only solidifies my position.