A New Year’s Day explosion at the Trump Hotel in Las Vegas has led to serious discussions about the role of generative AI after local police released key investigation details.

Matthew Livelsberger, a US Army soldier, is the main suspect in the case. Investigations revealed a ‘possible manifesto’ on his phone, along with an email intended for a podcaster and several letters.

Footage showed Livelsberger preparing for the event by pouring fuel over a truck before heading to the hotel site.

The Las Vegas Metro Police disclosed that the suspect made several inquiries to ChatGPT about explosives, how to detonate them using a gunshot, and legally acquiring guns and explosive materials.

Liz Bourgeois, a spokesperson for OpenAI, expressed sorrow over the incident, emphasizing the company’s commitment to responsible AI use.

According to Bourgeois, the AI tool provided information that was already public, along with warnings against harmful or illegal activities.

OpenAI is cooperating with law enforcement to assist with the ongoing investigation.

While officials are still investigating the cause of the explosion, one theory is that the ignition of fuel vapor by a gunshot could have triggered the blast, leading to a larger explosion with fireworks and other materials inside the truck.

The use of AI by the suspect, combined with the ability of investigators to trace his digital inquiries, underscores the emerging challenges and discussions about AI chatbot safety and privacy in real-world scenarios.