Florida Ag Announces Investigation Into Openai Over Shooting That Allegedly Involved Chatgpt

Trending 11 hours ago
This illustration image shows nan AI (Artificial Intelligence) smartphone app ChatGPT surrounded by different AI App successful Vaasa, connected June 6, 2023.Image Credits:Olivier Morin/AFP / Getty Images

1:11 PM PDT · April 9, 2026

Florida’s Attorney General, James Uthmeier, announced Thursday that his agency planned to analyse OpenAI complete nan alleged domiciled of ChatGPT successful a deadly shooting past year.

In April 2025, a gunman opened occurrence connected Florida State University’s campus, sidesplitting 2 and injuring five. Last week, attorneys for 1 of nan victims of nan shooting claimed that ChatGPT had been utilized to scheme nan attack. The family of nan unfortunate has said that they plan to writer OpenAI complete nan incident.

“AI should beforehand mankind, not destruct it,” Uthmeier said in a statement posted to X. “We’re demanding answers connected OpenAI’s activities that person wounded kids, endangered Americans, and facilitated nan caller FSU wide shooting. Wrongdoers must beryllium held accountable.” Uthmeier added successful a video that subpoenas were “forthcoming” arsenic portion of nan probe.

ChatGPT has been linked to a growing number of deaths and convulsive incidents — including murders, suicides, and shootings — and has added to worries complete nan advent of what psychologists call “AI psychosis,” delusions that are reinforced, encouraged, aliases deepened by communications pinch chatbots.

For example, Stein-Erik Soelberg, a man pinch a history of intelligence wellness issues, had regularly communicated pinch ChatGPT earlier he killed his mother and past himself past year, according to a Wall Street Journal investigation. The chatbot often seemed to reinforce nan paranoid thoughts that consumed him successful nan lead-up to nan murder-suicide.

When reached for remark by TechCrunch, an OpenAI spokesperson provided nan pursuing statement: “Each week, much than 900 cardinal group usage ChatGPT to amended their regular lives done uses specified arsenic learning caller skills aliases navigating analyzable healthcare systems. Our ongoing information activity continues to play an important domiciled successful delivering these benefits to mundane people, arsenic good arsenic supporting technological investigation and discovery. We build ChatGPT to understand people’s intent and respond successful a safe and due way, and we proceed improving our technology. We will cooperate pinch nan Attorney General’s investigation.”

Florida’s probe continues a drawstring of bad luck for OpenAI. A New Yorker profile connected Sam Altman published earlier this week showed disapproval and discontent wrong nan institution and among its investors, moreover quoting a Microsoft executive as saying: “I deliberation there’s a mini but existent chance he’s yet remembered arsenic a Bernie Madoff- aliases Sam Bankman-Fried-level scammer.” Meanwhile, a Stargate-related task successful nan United Kingdom had to beryllium paused, reportedly owed to precocious power costs and regulation.

Techcrunch event

San Francisco, CA | October 13-15, 2026

Lucas is simply a elder writer astatine TechCrunch, wherever he covers artificial intelligence, user tech, and startups. He antecedently covered AI and cybersecurity astatine Gizmodo. You tin interaction Lucas by emailing lucas.ropek@techcrunch.com.

More