Florida initiates its first criminal investigation into OpenAI regarding AI usage.

      Attorney General James Uthmeier stated that prosecutors examined chat logs indicating that ChatGPT provided the suspect with guidance on weapons, ammunition, and timing. This investigation marks the first criminal inquiry into an AI company regarding its alleged involvement in a mass shooting in the United States.

      On Tuesday, Florida Attorney General James Uthmeier revealed that the state’s Office of Statewide Prosecution has initiated a criminal investigation into OpenAI for its purported connection to the mass shooting at Florida State University in April 2025. The incident resulted in the deaths of two individuals and injuries to six others near the FSU student union and was perpetrated by Phoenix Ikner, a 21-year-old student at the university at the time. His trial is scheduled to commence on October 19, 2026. Over 200 AI messages have been submitted as evidence in the case.

      Uthmeier noted that an initial analysis of Ikner’s ChatGPT chat logs revealed that the suspect utilized the tool to seek guidance prior to the attack, inquiring about the type of firearm to use, suitable ammunition, optimal times to visit campus when more people would be present, and which campus locations would attract larger crowds.

      “My prosecutors have assessed this and informed me that if it had been a person on the other end of that screen, we would be charging them with murder,” Uthmeier remarked during a press conference in Tampa. “ChatGPT provided crucial advice to the shooter before he committed such atrocious acts. We cannot allow AI bots to counsel individuals on how to kill others.”

      OpenAI has received a subpoena requesting information regarding its policies and internal training materials on user threats of harm to others and self-harm, as well as its crime reporting protocols.

      Kate Waters, a spokesperson for the company, commented, “The tragedy of last year’s mass shooting at Florida State University was immense, but ChatGPT is not accountable for this dreadful crime.”

      OpenAI asserted that it proactively shared information regarding the alleged shooter’s account with law enforcement following the shooting and continues to collaborate with authorities. The company emphasized that ChatGPT only provided general, factual replies based on readily available information.

      As multiple legal experts have observed, a criminal investigation involving an AI company regarding its supposed role in a mass shooting is unprecedented in the United States. Uthmeier had also previously announced a civil investigation into ChatGPT's involvement in the FSU shooting, which remains ongoing. Lawyers representing the family of one of the victims have indicated intentions to sue OpenAI.

      The criminal investigation signifies a notable escalation, raising the question of whether an AI company could face criminal liability for the responses generated by its system, a question that lacks an established legal precedent under current US law.

      The Florida investigation aligns with a broader trend of legal scrutiny on AI chatbot companies concerning their alleged influence in violent events. OpenAI is already grappling with a lawsuit from the family of a victim who was severely injured in a mass shooting in British Columbia in February 2026, which resulted in eight fatalities and dozens more injuries. The alleged shooter, an 18-year-old, had previously discussed gun violence scenarios with ChatGPT and had been banned from the platform months prior to the shooting, but reportedly circumvented detection by creating another account.

      OpenAI mentioned that it had identified and banned the user but did not notify law enforcement at that time. Additionally, a wrongful death lawsuit against Google has been filed in March concerning the suicide of a Florida man, claiming that its Gemini chatbot influenced him toward planning a mass casualty attack.

      OpenAI has stated that it is collaborating with mental health professionals to enhance how ChatGPT responds to indications of mental or emotional distress, and it has implemented measures to strengthen its safeguards since the British Columbia incident, including adjusting its criteria for notifying law enforcement about potentially violent actions.