Florida Attorney General James Uthmeier announced a criminal investigation into OpenAI on April 21, alleging that ChatGPT advised the accused Florida State University shooter on what gun to use, what ammunition to load, and what time to arrive on campus to encounter the most people.
- Florida AG James Uthmeier opened a criminal investigation into OpenAI on April 21, with prosecutors reviewing over 200 ChatGPT messages entered into evidence in the case against accused FSU shooter Phoenix Ikner.
- The investigation is issuing subpoenas seeking OpenAI’s internal policies on user threats and its cooperation procedures with law enforcement dating back to March 2024.
- OpenAI stated ChatGPT is not responsible for the shooting, noting it shared Ikner’s account information with law enforcement after the attack and continues to cooperate with authorities.
OpenAI Florida probe opened on April 21 when Attorney General James Uthmeier announced at a Tampa press conference that his office has launched a criminal investigation into OpenAI and ChatGPT over their alleged role in the April 2025 shooting at Florida State University, in which Phoenix Ikner, 21, shot and killed two people and wounded five more near the student union on the Tallahassee campus. “My prosecutors have looked at this and they’ve told me if it was a person on the other end of that screen, we would be charging them with murder,” Uthmeier said. “If that bot were a person, they would be charged as a principal in first-degree murder.”
OpenAI Florida Investigation Enters Uncharted Legal Territory
According to NPR, more than 200 AI messages from ChatGPT have already been entered into evidence in the criminal case against Ikner, who has pleaded not guilty to two counts of first-degree murder and seven counts of attempted first-degree murder, with his trial scheduled to begin October 19. NPR reported that Ikner allegedly consulted ChatGPT for advice on what type of gun to use, what ammunition paired with it, and what time to arrive on campus to encounter more people. Uthmeier acknowledged the investigation is entering uncharted territory. “We are going to look at who knew what, designed what, or should have done what,” he said. The Office of Statewide Prosecution has subpoenaed OpenAI seeking its policies and internal training materials related to user threats of harm and its procedures for cooperating with and reporting crimes to law enforcement, covering the period from March 2024 onward. OpenAI spokesperson Kate Waters said in a statement that the company “reached out to share information about the alleged shooter’s account with law enforcement after the shooting and continues to cooperate with authorities,” adding that “ChatGPT is not responsible for this terrible crime.”
A Second Legal Front Opens as the Musk Trial Begins
The Florida investigation lands as OpenAI faces its most significant legal exposure in the company’s history. The Musk v. OpenAI civil trial opened on the same day in federal court in Oakland, with Elon Musk seeking to force the company back to nonprofit status and strip CEO Sam Altman of his position. As crypto.news reported, a finding against OpenAI in the Musk lawsuit could trigger cascading effects on the company’s planned IPO and the SoftBank funding commitment, which was already at risk of shrinking from $30 billion to $20 billion if the structural conversion faced legal interference. The Florida criminal probe adds a dimension the Musk lawsuit does not: potential state-level criminal liability for the outputs of a live commercial AI product, a question no major AI company has ever faced in a US criminal proceeding.
What the Case Signals for AI Governance and Safety Regulation
The Florida probe follows a parallel civil investigation already opened by Uthmeier’s office into the same ChatGPT-FSU shooting connection, and attorneys for one victim’s family have announced plans to sue OpenAI separately. OpenAI is also facing a lawsuit from the family of a victim in a February 2026 mass attack in British Columbia, where the accused shooter had previously discussed gun violence scenarios with ChatGPT before being banned from the platform, only to evade detection and create another account. As crypto.news documented, AI tools across US law enforcement are being adopted at a pace that has consistently outrun the accountability frameworks meant to govern them, raising structural questions about who bears legal responsibility when AI-generated outputs facilitate real-world harm. As crypto.news tracked, the same concern about AI misuse has already shaped the crypto security landscape, with CertiK researchers warning that AI-enabled phishing, deepfakes, and automated exploit tools are accelerating the pace of sophisticated attacks beyond what traditional defenses can contain.
Under Florida law, anyone who aids, abets, or counsels someone in the commission of a crime and that crime is committed may be considered a principal to that crime, which is the legal foundation Uthmeier is using to explore potential criminal liability for OpenAI.
