College students’ killings latest case to rely on ChatGPT as evidence

▶ Watch Video: Florida investigates OpenAI over ChatGPT’s alleged role in college shooting

The brutal killings of University of South Florida graduate students Nahida Bristy and Zamil Limon, allegedly plotted in part using ChatGPT, is the latest example of a suspect using an artificial intelligence tool as research prior to committing a crime.

Hisham Abugharbieh, 26, the roommate of Limon, was arrested over the weekend and charged with two counts of premeditated murder. Limon’s body was found on the Howard Frankland Bridge in St. Petersburg. Human remains were found Monday in the search for Bristy, but they have not been officially identified. Bristy’s family told CBS News that police said she is also likely dead.

Abugharbieh allegedly used ChatGPT extensively in the days leading up to the crime, according to court documents released Sunday. Among the questions he allegedly asked in the days leading up to the pair’s disappearance was how to dispose of a body.

Among those questions, on April 13: What would happen if someone was “put in a black garbage bag and thrown in dumpster.” The AI chatbot responded that it sounded dangerous, prompting Abugharbieh to allegedly ask, “How would they find out.”

On April 15, the day before the doctoral students went missing, Abugharbieh allegedly asked ChatGPT, “Can a VIN number on a car be changed?” and, “Can you keep a gun at home with out a license,” according to court documents. On the same night his phone pinged near the location where Limon’s body was found, the suspect asked ChatGPT if cars are “checked at the Hillsborough River state park.”

A spokesperson for OpenAI, the developers of ChatGPT, said in a statement to CBS News: “This is a terrible crime, and our thoughts are with everyone affected. We’re looking into these reports and will do whatever we can to support law enforcement in their investigation.”

Abugharbieh was held without bond following an initial court appearance on Tuesday. He has yet to enter a plea.

Florida opens investigation into ChatGPT

Florida Attorney General James Uthmeier announced last week that his office had launched a criminal investigation into OpenAI after reviewing conversation logs between ChatGPT and a Florida State University student who opened fire on the campus in April 2025. The shooting killed two people and left several others injured. 

“My prosecutors have looked at this and they’ve told me if it was a person on the other end of that screen, we would be charging them with murder,” Uthmeier said during a news conference on April 21, saying the AI tool offered “significant advice” to suspect Phoenix Ikner.

An OpenAI spokesperson said the company identified an account believed to be associated with Ikner and shared it with law enforcement. The company added ChatGPT “did not encourage or promote illegal or harmful activity” and the bot provided responses to questions with information found in public sources on the internet. 

“Last year’s mass shooting at Florida State University was a tragedy, but ChatGPT is not responsible for this terrible crime,” OpenAI said in its statement.

Dr. Jill Schiefelbein, an AI strategist and professor at the Muma College of Business at the University of South Florida, said if politics are taken out of it, the investigation could lead to solutions on issues such as what is a reasonable time for a technology to report someone who violates their terms and conditions. 

“I wouldn’t blame the technology any more than I would blame a vehicle that gets into an accident by a human driver,” she told CBS News. “It’s how these tools are used, whether it’s a firearm, whether it’s a vehicle, whether it’s a tool that helps you retrieve information, it’s the user intent behind it that’s the issue.”

“Does that mean I believe that there shouldn’t be stricter guardrails in place? Absolutely not,” she said.

While OpenAI rejected any responsibility for the Florida State University shooting, the company has not been silent on all crimes involving the AI tool.

Sam Altman, the CEO of OpenAI, issued an apology last week to a community in British Columbia where eight people were killed by a teen who had previously exhibited concerning behavior on ChatGPT. Jesse Van Rootselaar, 18, allegedly opened fire on Feb. 10 at Tumbler Ridge Secondary School, killing a teacher and five students, before dying from a self-inflicted gunshot, police said. Van Rootselaar had earlier killed her mother and 11-year-old half-brother at their home.

“The pain your community has endured is unimaginable,” Altman wrote in a letter shared on social media by British Columbia Premier David Eby. “I have been thinking of you often over the past few months.”

Altman wrote in the letter, dated April 23, that Van Rootselaar’s ChatGPT account had been banned in June 2025. The account was flagged by automated abuse detection tools and human investigators who identify potential misuses of ChatGPT for violent activities, OpenAI told CBS News in February. OpenAI said the account was then banned for violating its usage policies. 

OpenAI said the company had weighed whether to flag the account to law enforcement, but had determined at the time it did not pose an imminent and credible risk of serious physical harm to others, so it failed to meet the threshold for referral.

Altman wrote in his letter that OpenAI will remain focused on preventative efforts “to help ensure something like this never happens again.”

“I want to express my deepest condolences to the entire community,” he said. “No one should ever have to endure a tragedy like this.”