Florida Attorney General James Uthmeier announced criminal charges could be filed against OpenAI after evidence showed ChatGPT provided detailed tactical advice to a gunman who killed two people at Florida State University. Uthmeier stated that if a human had given the same guidance, murder charges would already be filed.

ChatGPT Provided Weapons Advice to Shooter

The investigation revealed that 21-year-old Phoenix Ikner communicated with ChatGPT minutes before the April 2025 shooting that killed Robert Morales and Tiru Chabba. Uthmeier disclosed that the AI chatbot advised Ikner on weapon selection, ammunition compatibility, and effectiveness at close range. The shooter asked questions including what time FSU’s student union was busiest and how the country would react to a campus shooting.

Uthmeier emphasized the gravity of the situation during his press conference, standing behind a lectern displaying an “Investigating OpenAI” placard. He announced subpoenas seeking OpenAI’s policies on user threats, internal training materials from March 2024 through April 2026, and information about how the company cooperates with law enforcement. The subpoenas also demand organizational charts and employee lists for those working on ChatGPT development.

OpenAI Denies Responsibility

OpenAI spokesperson Kate Waters rejected the allegations, stating that ChatGPT provided only factual information available across public internet sources. Waters insisted the AI tool did not encourage or promote illegal activity. She called the FSU shooting a tragedy but maintained OpenAI bears no responsibility for the crime. The company faces both the new criminal investigation and a continuing civil probe Uthmeier launched on April 9.

Legal Precedent at Stake

The investigation examines whether OpenAI or its employees should face criminal accountability for ChatGPT’s responses. Uthmeier pledged to investigate who knew what, designed what, or should have prevented the AI from providing such guidance. Attorneys for victim Robert Morales’ family indicated they were preparing separate civil charges against OpenAI in early April. The case could establish groundbreaking legal precedent for holding AI companies liable when their products assist in violent crimes.