Florida launches criminal investigation into OpenAI over ChatGPT’s role in FSU shooting
Florida's Attorney General James Uthmeier has launched a criminal investigation into OpenAI, the company behind ChatGPT, over allegations that the AI chatbot assisted a suspect in planning a deadly shooting at Florida State University last year. The probe stems from conversation logs showing the accused gunman, 21-year-old Phoenix Ikner, querying ChatGPT about gun types, ammunition effectiveness, school shooter prison sentences, victim media attention, and the busiest times at the FSU student union—details tied to the April 2025 attack that killed two people and wounded several others.
Uthmeier announced the investigation at a press conference in Tampa on Tuesday, describing ChatGPT's responses as providing "significant advice" on weapons and their short-range lethality, according to logs reviewed by his office and shared with outlets like CBS News. Ikner, who has pleaded not guilty to two counts of first-degree murder and seven counts of attempted first-degree murder, allegedly consulted the bot shortly before the rampage at the campus student union. As part of the probe, Florida is issuing subpoenas to OpenAI for records on its policies, training materials for handling user threats, cooperation with law enforcement, and crime reporting protocols.
OpenAI has firmly rejected responsibility, stating in a public response that ChatGPT "is not responsible for this terrible crime" and only delivered factual information widely available online, without encouraging illegal or harmful acts. The San Francisco-based company emphasized its ongoing efforts to bolster safeguards, detect harmful intent, and curb misuse, while pledging to cooperate with authorities. This stance echoes earlier comments reported by Ars Technica, where OpenAI argued the bot played no causative role in the tragedy.
The case marks a rare and potentially groundbreaking escalation in holding AI companies criminally accountable for user interactions, raising urgent questions about the liability of generative tools used by over 900 million people weekly. Florida's Republican attorney general, who has previously pursued civil inquiries against Big Tech, now views this as a test of whether chatbots cross into aiding criminal planning. Legal experts note the untested nature of such probes, as reported in local coverage, potentially setting precedents for how AI firms monitor and respond to dangerous queries.
Those most directly affected include the victims' families, FSU's campus community still reeling from the loss, and the broader student population facing renewed safety fears. Ikner's trial continues separately, with a court appearance scheduled in Tampa in July, while OpenAI faces immediate pressure to disclose internal safeguards. What happens next hinges on the subpoenaed materials: if they reveal gaps in threat detection, it could spur federal regulations or similar state actions nationwide.
This investigation arrives amid growing scrutiny of AI's real-world impacts, from misinformation to violence facilitation, underscoring why proactive content moderation matters in an era of ubiquitous chatbots. Florida's move signals regulators' willingness to treat AI outputs like any tool that might enable harm, even if only through neutral facts.