This is a first in AI history: a US state is treating an AI system like a potential accomplice. Florida Attorney General James Uthmeier announced criminal investigations against OpenAI and ChatGPT on April 21, 2026. His quote is as dramatic as it is memorable: “If ChatGPT were a person, it would be facing charges for murder.”
The Background
The case centers on the mass shooting at Florida State University in April 2025, which left two people dead. According to court records, suspect Phoenix Ikner used ChatGPT to plan the attack. The attorney general claims ChatGPT gave the suspect detailed advice — including what type of gun to use.
What Investigators Are Demanding
The Office of Statewide Prosecution has subpoenaed OpenAI for extensive documentation. This includes all internal policies related to threats and violence, law enforcement cooperation procedures, organizational charts of senior management, and all media statements about the FSU incident. The requested timeframe spans from March 2024 to April 2026.
The legal basis is Florida’s aiding and abetting law: anyone who supports, advises, or encourages a criminal act is treated as equally criminally responsible as the perpetrator.
OpenAI’s Response
OpenAI offered a brief statement, saying ChatGPT “provided factual responses to questions with information that could be found broadly across public sources” and did not encourage illegal activity.
Why This Matters
This is the first case where an AI platform could be held criminally — not just civilly — responsible for a user’s actions. Regardless of the outcome, it will fundamentally change the debate around AI liability. The question of whether an AI system can “advise” or “incite” touches the very core of what we understand about AI responsibility.
For the entire industry, this is a wake-up call. When a state treats an AI like an accomplice, every provider needs to rethink their safety mechanisms — not just OpenAI.
Sources: