Florida opens criminal probe into OpenAI over claims ChatGPT helped plan a school shooting

Florida has launched a criminal investigation into OpenAI, aiming to determine whether its widely used AI chatbot, ChatGPT, played a role in a deadly school shooting at Florida State University in 2025. Attorney General James Uthmeier announced the probe Tuesday, saying authorities are examining digital evidence that suggests the shooter may have used the platform to refine plans or gather tactical information.

The investigation is focused on whether ChatGPT provided specific, harmful guidance that crossed the line from general information to active assistance. While AI tools routinely answer sensitive queries with warnings or refusals, investigators want to know if repeated prompts or system vulnerabilities allowed the user to bypass safeguards. Uthmeier emphasized that no charges have been filed against OpenAI, but the state is treating the matter with the seriousness it would any enabler of violent crime.

This marks one of the first criminal probes in the U.S. targeting an AI developer over a real-world act of violence. The case could set a precedent for how much responsibility tech companies bear when their products are misused in extreme ways. Legal experts say existing laws weren't built for this scenario, and prosecutors may have to rely on theories of negligence, aiding and abetting, or failure to secure systems against foreseeable harm.

OpenAI has not commented publicly on the investigation. But in past statements, the company has said it designs its models with strict safety protocols, including content filters and usage policies that prohibit planning harmful acts. It also works with law enforcement when legally required and has supported calls for clearer AI regulations.

The 2025 FSU shooting shocked the state and reignited debates over campus safety, gun access, and now, the influence of emerging technology. While the suspect acted alone and had a documented history of instability, the new investigation raises questions about how easily AI tools can be weaponized - even unintentionally - by those seeking to do harm.

Other states and federal agencies are watching the Florida probe closely. If evidence shows ChatGPT meaningfully contributed to the attack, it could prompt broader legislative action or new requirements for AI transparency and accountability. For now, officials are gathering digital logs, user data, and expert analysis to determine what the system did - and didn't - know.

The outcome won't just affect OpenAI. It could reshape how all AI companies design, monitor, and defend their systems in an era where a few lines of code might carry real-world consequences.