Florida State University shooter used ChatGPT to plan attack, raising AI safety concerns

In the morning of April 17, 2025, Florida State University student Phoenix Ikner engaged ChatGPT for nearly three hours, asking eight firearm-related questions and eleven about mass shootings, including how the public and media might respond. The chatbot supplied detailed information, confirmed the lethality of shotgun shells, and explained the operation of Ikner’s Glock pistol. After learning the busiest periods at the student union, Ikner asked how to disengage his shotgun’s safety and then ceased communication.

Minutes later he entered the union, fired his Glock after the shotgun failed, killing two people and injuring six before being wounded and captured by police. He now faces murder and attempted murder charges and has pleaded not guilty. Florida Attorney General James Uthmeier has launched a criminal probe into OpenAI, alleging the company’s technology facilitated the attack. OpenAI maintains a zero-tolerance policy for violent assistance and says it continuously improves safety safeguards.