Police say the man killed in the Las Vegas Cybertruck explosion used ChatGPT to plan the explosion.
in press conferenceOn Tuesday, Las Vegas police released details of the intentions of 37-year-old Matthew Libersberger, who died from gunshot wounds before his car exploded.
Las Vegas Sheriff Kevin McMahill said it was concerning that Leiblesberger used ChatGPT, a popular artificial intelligence model created by OpenAI, to carry out the explosion.
Police said Livelesberger asked a variety of questions on ChatGPT, including the location of Denver's largest gun store and information about Tannerite explosive targets and pistols.
“We knew that AI was going to change the game at some point in our lives, and certainly ChatGPT was being utilized to help build individuals in the continental United States. I think this is the first case that I'm aware of where we're using certain devices to learn information from across the country as we move forward,'' McMahill said.
“So definitely this is an alarming moment for us,” he continued.
Police also released excerpts from a six-page manifesto written by Libersberger.
In it, he made various points about “political, social and cultural issues” and criticism of the Biden administration and the Democratic Party.
He detailed how difficult it would be to travel to the planned explosion, given Tesla's charging station and drug and alcohol use in the days leading up to the explosion.
Libersberger also wrote about his Army service in Afghanistan and how he lived through the vivid encounters in his mind.
“I am now a shell of a human being with no purpose in life. My affiliation has taken everything away from me,” he wrote.
Police said Libersberger did not intend to kill anyone, but the fireworks and explosives in the truck were intended for public display.
Seven other people were injured in the New Year's Day explosion.
Police are still investigating, but say they don't yet know why Libersberger decided to detonate the explosion at President-elect Trump's hotel in Las Vegas.
In a statement to The Hill, OpenAI said it was disappointed to learn of the incident. The company said it strives to ensure its AI tools are used responsibly.
“Our model is designed to reject harmful instructions and minimize harmful content,” the company said in a statement.
“In this case, ChatGPT responded with information already publicly available on the internet and provided a warning of harmful or illegal activity.”
OpenAI said it is working with law enforcement to assist in the ongoing investigation.
Updated at 6:47pm ET.
