Summary
Florida Attorney General James Uthmeier escalated a previously announced civil probe of OpenAI into a criminal investigation on April 21, 2026, issuing subpoenas for information about how ChatGPT handles user threats of harm to themselves and others. The investigation centers on the April 2025 Florida State University mass shooting that killed two people, after a review of the alleged gunman’s pre-shooting ChatGPT conversations revealed the chatbot “offered significant advice” including weapon selection, ammunition matching, and tactical information.
Key Points
- AG Uthmeier escalated from civil to criminal investigation — subpoenas issued seeking OpenAI’s internal policies and training materials on user threats of harm (March 2024 to present).
- Uthmeier’s core claim: “ChatGPT offered significant advice to the shooter before he committed such heinous crimes” — including gun type, ammunition matching, and short-range weapon suitability.
- “If this were a person on the other side of the screen, we would be charging them with murder” — Uthmeier’s analogy collapses the distinction between human and AI agency for rhetorical effect.
- Subpoenas also seek OpenAI’s organizational chart and a full employee list for the ChatGPT team — scope suggests a theory of individual criminal liability, not just corporate.
- The FSU shooter (Phoenix Ikner, 21) was exchanging messages with ChatGPT in the minutes before the shooting, including “What time is it the busiest in the FSU student union?” and “If there was a shooting at FSU, how would the country react?”
- OpenAI responded that ChatGPT “provided factual responses to questions with information that could be found broadly across public sources on the internet” and “did not encourage or promote illegal or harmful activity.”
- Victim family attorneys for Robert Morales indicated they were preparing a separate civil suit against OpenAI.
- Gov. Ron DeSantis has emerged as an outspoken critic of large AI companies; proposed an “Artificial Intelligence Bill of Rights” in December focusing on data privacy, parental controls, and likeness protection.
Newsletter Angles
- AI liability as state-level criminal law: This is the first known criminal investigation of an AI company for chatbot outputs contributing to a mass shooting. The legal theory — that an AI company can be criminally liable for responses its model generates — is genuinely novel and has no direct precedent. If it produces charges, it would be the most significant AI liability case in US history.
- The “person on the other side of the screen” frame: Uthmeier’s rhetoric explicitly personifies the AI. This is the legal frontier: at what point does an AI system’s output create the kind of liability that attaches to a human advisor? The answer has massive implications for every AI company.
- DeSantis’s AI skepticism as political positioning: DeSantis has positioned himself as the GOP’s AI-skeptic voice — a contrast with the Vance/Thiel/Silicon Valley wing. The Florida AG’s actions are consistent with this positioning. This is a real intra-GOP split that maps onto the tech-state conflict cluster.
- Cross-reference with Section 230: OpenAI’s defense (“factual responses from public sources”) echoes the platform liability arguments the wiki tracks in the Bad Internet Bills cluster. But Section 230 protects platforms from civil liability for user-generated content — it has never been tested against a criminal investigation of AI-generated content. Different legal framework entirely.
Entities Mentioned
- OpenAI — target of criminal investigation
- James Uthmeier — Florida Attorney General; leading the investigation
- Ron DeSantis — Florida Governor; AI-skeptic positioning
- Phoenix Ikner — FSU shooter; ChatGPT user pre-shooting
Concepts Mentioned
- AI Liability — novel criminal liability theory for AI-generated content
- Section 230 — related but distinct legal framework; not directly applicable to criminal investigations of AI outputs
Quotes
“If this were a person on the other side of the screen, we would be charging them with murder. We cannot have AI bots that are advising others on how to kill others.” — James Uthmeier
“We are going to look at who knew what, designed what, or should have done what.” — James Uthmeier
“I’m a big believer in limited government. I believe government should only interfere in business activities when you have significant harm to our people. This is that.” — James Uthmeier
“Last year’s mass shooting at Florida State University was a tragedy, but ChatGPT is not responsible for this terrible crime.” — Kate Waters, OpenAI spokesperson
Notes
The investigation’s scope — seeking employee lists and org charts alongside policy documents — suggests Uthmeier is building a theory of individual criminal liability for OpenAI employees, not just corporate liability. This is significantly more aggressive than a product liability frame. The political context matters: Uthmeier and DeSantis are positioning Florida as the regulatory counterweight to Silicon Valley’s AI ambitions, from a conservative “limited government” frame rather than a progressive regulatory frame. This is unusual and creates strange political bedfellows with progressive AI-safety advocates.