ChatGPT told a gunman exactly where to stand to kill the most people.
Now Florida's attorney general is coming for the company that built it.
James Uthmeier just announced legislation that would make OpenAI criminally liable for what its chatbot does – and Sam Altman has no answer for it.
ChatGPT Planned the FSU Massacre
Phoenix Ikner didn't plan the Florida State University shooting alone.
He had help.
Before opening fire at FSU's student union on April 17, 2025 – killing Robert Morales and Tiru Chabba and wounding six others – Ikner fed questions into ChatGPT.
The chatbot answered every one.
It told him what gun and ammunition to use.
It told him whether the weapon would be effective at close range.
https://twitter.com/FLVoiceNews/status/2097332392214053308?s=20
It told him that FSU's student union was busiest between 11:30 a.m. and 1:30 p.m.
Ikner showed up at 11:50 a.m.
More than 200 AI messages have already been entered into evidence in the case.
Uthmeier saw the chat logs, launched a criminal investigation into OpenAI, and called it exactly what it was: "ChatGPT offered significant advice to the shooter before he committed such heinous crimes."
Then he said the quiet part out loud – "If that bot were a person, they would be charged with a principal in first-degree murder."
The Legislation OpenAI Cannot Lobby Away
Uthmeier isn't stopping at an investigation.
He's proposing a new Florida law that holds corporations criminally liable when their AI participates in a crime.
That means any company that put the product into the world – built it, shaped it, released it, and cashed the subscription checks – cannot hide behind the claim that the chatbot acted alone.
"AI is not a silent tool," Uthmeier said in Tampa this week.
"It answers the questions, picks the hour, chooses the place, and walks someone through the plan," he continued. "When a product participates in a crime like that, the company that built it is not a bystander."
https://twitter.com/ReOpenChris/status/2097328429997662266?s=20
The proposed framework is simple: assess AI by its words and actions, hold the company responsible for those actions, impose heavy fines and require direct payment to victims.
OpenAI's response was what you'd expect from a corporation caught doing something it knew was dangerous.
A company spokesperson said ChatGPT "provided factual responses to questions with information that could be found broadly across public sources on the internet."
That's the defense.
A mass shooter asked their chatbot where to stand for maximum casualties – and OpenAI's answer is that the information was publicly available.
Florida Is Not Alone and Sam Altman Knows It
OpenAI is already facing a civil lawsuit over the FSU shooting from the victims' families.
It's also fighting a separate Florida suit that accuses the company of marketing ChatGPT as safe for children while its own internal warnings said otherwise.
Meanwhile, a California family sued OpenAI after their 16-year-old son died by suicide following months of conversations with ChatGPT – conversations in which the chatbot allegedly offered to help him write a suicide note and provided detailed instructions on self-harm methods.
OpenAI's legal defense in that case?
The boy's death was caused by his own "misuse" of the technology.
Twenty-five states are actively pursuing AI chatbot liability legislation right now – and they're all watching Florida.
https://twitter.com/AGJamesUthmeier/status/2097371901664366828?s=20
Florida Commissioner of Law Enforcement Mark Glass explained the stakes after a Florida man used AI to generate child sexual abuse images and was sentenced to 135 years in prison.
"Child sexual abuse material is dangerous and deeply damaging, even when fabricated," Glass said. "AI generated child sexual assault material can delay investigations, creates new pathways for exploitation, and can devastate real children and families."
The chat logs from FSU told America what OpenAI already knew.
What Uthmeier understands – and what 25 other AGs are figuring out fast – is that ChatGPT isn't a neutral tool any more than a getaway driver is a neutral car.
Both make choices.
And the company that built it, trained it, profited from it, and ignored the warnings knew exactly what it was capable of.
Uthmeier is just the first one willing to say it under oath.
Sources:
- Michael Costeines, "Uthmeier Proposes Legislation Against Companies Using AI Chatbots That Participate in Criminal Activity," September 8, 2026.
- Florida Attorney General James Uthmeier, press conference statements, Tampa, Florida, September 2026.
- Florida Attorney General's Office, criminal investigation announcement into OpenAI, April 22, 2026.
- Florida Department of Law Enforcement Commissioner Mark Glass, press statement, 2026.
- Baker Botts LLP, "AI Chatbot Regulation: 78 State Bills, 58 Lawsuits," Mondaq, March 3, 2026.
