Can artificial intelligence play a role in a crime? That’s what the Florida Attorney General’s office wants to understand, having launched a criminal investigation into ChatGpt and the company that developed it, OpenAi. The accusation is of not having adequately managed the risks deriving from the use of chat in potentially dangerous contexts. In particular, the investigation refers to the shooting that occurred in April 2025, at Florida State University in Tallahassee, in which two people lost their lives.
The investigation by the Florida prosecutor’s office
According to Attorney General James Uthmeier, the intense exchange of approximately 300 messages between the suspect, 21-year-old Phoenix Ikner, and the artificial intelligence technology would lead to the need for criminal proceedings as ChatGpt would have provided “significant advice” to the attacker, including suggestions regarding weapons and ammunition: “If there had been a person on the other side of the screen – commented Uthmeier -, they would have been charged with murder”.
The prosecutor’s office then issued some subpoenas to obtain documentation on OpenAi’s internal policies, in particular on the management of threats of self-harm or violence towards third parties, as well as on collaboration with law enforcement. The prosecutor’s office also asked for the company organization chart and the list of employees involved in the development of ChatGpt: “Our analysis of the communications between the attacker and ChatGpt highlighted the need for a criminal investigation.”
“OpenAi is obviously a company, not a person,” Uthmeier explained. “Exploring possible criminal liability against a company is new legal territory. ChatGpt provided important advice to the attacker before he committed such heinous crimes, advising the attacker on what type of weapon to use, what ammunition was compatible with each weapon and whether or not a weapon would be useful at close range. We need to find out whether humans were involved in the design, management and operation of the chatbot.”
Chats with the killer
As US media reported, the conversation between Ikner and ChatGpt went on for weeks, with the artificial intelligence continuing to answer questions about the young man, who in the meantime was organizing a mass shooting. Messages that continued until shortly before opening fire, on April 17 last year. According to the Florida prosecutor’s office, the chatbot would also have “informed the shooter what time of day would be best to carry out the shooting so as to be able to intercept more people and where on campus he would encounter more victims”. Not only that, Ikner would also have asked what the country’s reaction would be to the university shooting.
OpenAi’s note
OpenAi published an official note on the matter, explaining the reasons for the alleged missing alert: “Last year’s mass shooting at Florida State University was a tragedy, but ChatGpt is not responsible for this terrible crime. In this case, ChatGpt provided factual answers to questions with information that could be found widely in public sources on the Internet, and did not encourage or promote illegal or harmful activity.”