Cybertruck Bomber Used ChatGPT to Plan His Attack

A weird new development.

TerrorGPT

Las Vegas police officials have revealed that active duty Green Beret soldier Matthew Livelsberger, who fatally shot himself and blew up a rented Tesla Cybertruck just outside the Trump Towers hotel in Las Vegas last week, used generative AI — including OpenAI’s ChatGPT — to plan his attack.

As the Associated Press reports, a review of Livelsberger’s queries on ChatGPT suggests he was not only researching explosive targets, but was also studying the speed of bullets, how far they would travel, and even whether fireworks were legal in Arizona (they are, but only 27 days of the year).

It’s a weird new development — and one that shows, yet again, that every new technology is eventually subsumed into the darkest corners of society for violent and twisted purposes.

“This is the first incident that I’m aware of on US soil where ChatGPT is utilized to help an individual build a particular device,” said Las Vegas Metropolitan Police Department sheriff Kevin McMahill during a press conference. “It’s a concerning moment.”

Wake Up

The Afghanistan veteran shot himself with a handgun before blowing up the Cybertruck using improvised explosives and fireworks. He didn’t take the lives of any bystanders, but injured seven people.

As far as a motive is concerned, Livelsberger likely was “struggling with PTSD and other issues,” as FBI special agent Spencer Evans told reporters.

Livelsberger also made bizarre comments about “Chinese” drones with “gravitic propulsion systems” in a widely circulating email that officials believe was written by him.

He also wrote in a note that the US was “terminally ill and headed toward collapse” and that he intended the attack to be a “wake up call.”

He also vowed to “cleanse” his mind “of the brothers I’ve lost and relieve myself of the burden of the lives I took.”

Interestingly, he wrote in his note that he had no ill will towards Tesla CEO Elon Musk and that the US needs to “rally around” the entrepreneur and president-elect Donald Trump.

Apart from plenty of evidence suggesting Livelsberger was a deeply troubled individual, his use of ChatGPT stands out. It’s unclear whether he managed to get the tool to spit out information on specifically illegal activities, such as how to build an explosive. But given the sorry state of guardrails and how easily they’re circumvented, it’s certainly not impossible — so we’ll be watching with interest.

More on the attack: The Cybertruck Bomber Seems to Have Made Strange Claims About the Mysterious Drone Sightings