Las Vegas police release ChatGPT logs from the suspect in the Cybertruck explosion

They confirmed that the suspect, an progressive work worker successful the US Army named Matthew Livelsberger, had a “possible manifesto” saved connected his phone, successful summation to an email to a podcaster and different letters. They besides showed video grounds of him preparing for the detonation by pouring substance onto the motortruck while stopped earlier driving to the hotel. He’d besides kept a log of expected surveillance, though the officials said he did not person a criminal grounds and was not being surveilled aliases investigated.

The Las Vegas Metro Police also released respective slides showing questions he’d posed to ChatGPT respective days earlier the explosion, asking astir explosives, really to detonate them, and really to detonate them pinch a gunshot, arsenic good arsenic accusation astir wherever to bargain guns, explosive material, and fireworks legally on his route.

Asked astir the queries, OpenAI spokesperson Liz Bourgeois said:

The officials opportunity they are still examining imaginable sources for the explosion, described arsenic a deflagration that traveled alternatively slow arsenic opposed to a precocious explosives detonation that would’ve moved faster and caused much damage. While investigators opportunity they haven’t ruled retired different possibilities for illustration an electrical short yet, an mentation that matches immoderate of the queries and the disposable grounds is that the muzzle flash of a gunshot ignited substance vapor/fireworks fuses wrong the truck, which past caused a larger detonation of fireworks and different explosive materials.

Trying the queries successful ChatGPT coming still works, however, the accusation he requested doesn’t look to beryllium restricted and could beryllium obtained by astir hunt methods. Still, the suspect’s usage of a generative AI instrumentality and the investigators’ expertise to way those requests and coming them arsenic grounds return questions astir AI chatbot guardrails, safety, and privateness out of the hypothetical realm and into our reality.