AP WASHINGTON

The highly decorated soldier who exploded a Tesla cybertruck outside the Trump hotel in Las Vegas used generative AI, including ChatGPT, to help plan the attack, Las Vegas police said Tuesday.

Nearly a week after 37-year-old Matthew Livelsberger fatally shot himself, officials said according to writings, he didn't intend to kill anyone else.

An investigation of Livelsberger's searches through ChatGPT indicate he was looking for information on explosive targets, the speed at which certain rounds of ammunition would travel and whether fireworks were legal in Arizona.

Kevin McMahill, sheriff of the Las Vegas Metropolitan Police Department, called the use of generative AI a "game-changer" and said the department was sharing information with other law enforcement agencies.

"This is the first incident that I'm aware of on US soil where ChatGPT is used to help an individual build a particular device," he said. "It's a concerning moment." 

In an emailed statement, OpenAI said it was committed to seeing its tools used "responsibly" and that they're designed to refuse harmful instructions.

"In this case, ChatGPT responded with information already publicly available on the internet and provided warnings against harmful or illegal activities. We're working with law enforcement to support their investigation," the emailed statement said.

Launched in 2022, ChatGPT is part of a broader set of technologies developed by the San Francisco-based startup OpenAI. Unlike previous iterations of so-called "large language models," the ChatGPT tool is available for free to anyone with an internet connection and designed to be more user-friendly.

Authorities also said they uncovered a six-page document that they have not yet released because they're working with Defence Department officials since some of the material could be classified. They added that they still have to review contents on a laptop, mobile phone and smartwatch.