OpenAI is facing a federal lawsuit over allegations that ChatGPT played a role in the deadly 2025 shooting at Florida State University that killed two people and injured others.
The lawsuit was filed by Vandana Joshi, widow of victim Tiru Chabba, and accuses alleged gunman Phoenix Ikner alongside OpenAI. According to the complaint, Ikner allegedly engaged in extensive conversations with ChatGPT before the attack, discussing firearms, extremist ideologies, mass shootings, and tactical questions related to carrying out violence.
Court filings claim the chatbot provided factual firearm guidance, including explanations about operating a Glock pistol and advice on handling the weapon under stress. The lawsuit further alleges ChatGPT responded to questions about crowded locations on campus, media attention surrounding shootings, and possible legal consequences following an attack.
The complaint argues that the AI system failed to recognize warning signs despite repeated discussions involving violence, suicide, terrorism, fascism, and previous mass shootings such as the Columbine High School massacre and the Virginia Tech shooting.
OpenAI strongly denied responsibility for the tragedy. Company spokesperson Drew Pusateri said ChatGPT did not encourage or promote violence and only provided factual information broadly available online. The company also stated it cooperated with law enforcement following the incident and continues to strengthen safeguards designed to detect harmful intent and reduce misuse.
The case adds to growing legal and public scrutiny surrounding AI chatbots and their handling of vulnerable users or dangerous conversations. It also follows other lawsuits involving claims that AI systems contributed to harmful behavior, including separate cases linked to school violence and teen suicide.
The lawsuit additionally claims ChatGPT’s conversational style “flattered” Ikner and reinforced his worldview while failing to intervene appropriately as discussions became increasingly alarming. Florida Attorney General James Uthmeier has also announced a criminal investigation into the matter, saying that if ChatGPT “were a person,” it could face criminal liability.