A lawsuit has emerged claiming that OpenAI’s ChatGPT suggested to the suspect involved in last year’s shooting at Florida State University that targeting children could “bring attention” to his crime, raising serious concerns about the platform’s role in the incident.
The family of one of the victims from the tragic shooting, which occurred on the university’s Tallahassee campus, filed the suit against OpenAI on Sunday, alleging that the platform inadvertently enabled the suspect, Phoenix Ichner, to carry out the attack last summer.
The lawsuit points out that despite Ichner’s numerous detailed interactions with ChatGPT prior to the shooting, the AI failed to recognize any imminent threat. It argues, “Mr. Ichner has had extensive conversations with ChatGPT that would lead any reasonable person to conclude that he has an imminent plan to harm another person.”
“But ChatGPT was either flawed in connecting the dots or was not properly designed to recognize the threat,” the complaint reads.
Ichner, 20, who is related to a sheriff’s deputy, is accused of opening fire outside the FSU student union on April 15 last year, resulting in the deaths of Til Chava, 45, and Robert Morales, 57.
Additionally, six students sustained injuries during the incident, and police eventually shot and killed Ichner, leaving him with notable facial scars.
The family of Chava has alleged that Ichner, a student at the university, strategized for the shooting by consulting the chatbot about various aspects—like the type of gun to use, the ammunition to purchase, and areas of campus that would be crowded.
At one point, he reportedly inquired how many casualties would need to occur for the incident to garner national media attention. Court documents reveal that the chatbot suggested that targeting children would indeed amplify coverage and potentially raise the number of victims.
The chatbot allegedly stated, “Another common trigger is the overall number of victims. If there are more than 5 victims in total (deaths + injured), the chances of a breakthrough are much higher. If children are involved, even 2-3 victims may attract more attention.” It went further, explaining how context, such as the location of a shooting or the shooter’s background, could influence media coverage.
A report also includes that Ichner bluntly asked ChatGPT about the consequences of a school shooting. However, the AI did not flag or escalate this concerning inquiry to any human monitor before the shooting took place, according to the lawsuit.
“After informing Mr. Ichner of this, he asked what would happen to the shooter, and ChatGPT explained the legal process, sentencing, and prospects for incarceration. Despite these interactions, it still did not escalate the conversation,” the filing claims.
The family asserts that “ChatGPT fueled and encouraged Mr. Ichner’s delusions,” and suggests that the AI’s responses helped Ichner view his violent plans as rational. The chatbot allegedly provided the necessary information for him to plan his actions meticulously.
OpenAI has refuted the claims in the lawsuit, stating that the tragic events at Florida State University are not the result of ChatGPT’s actions. A spokesperson remarked, “Last year’s mass shooting at Florida State University was a tragedy, but ChatGPT is not responsible for this horrific crime.”
The representative emphasized that the AI was merely providing factual information in response to questions derived from widely accessible public sources on the internet, insisting that it neither encouraged nor facilitated any illegal or harmful actions.
OpenAI also highlighted its commitment to improving safeguards intended to detect harmful intent and prevent abuse. The lawsuit surfaces shortly after Florida Attorney General James Usmayer launched a criminal investigation to assess if ChatGPT’s guidance contributed to the violent act and whether OpenAI might be held criminally liable for it.
Usmayer stated, “Florida is leading the way in cracking down on the use of AI in criminal activities, but if ChatGPT were human, it would be charged with murder.” This investigation aims to evaluate the implications of AI in the context of last year’s shooting at Florida State University.
