Artificial intelligence saved a life: ChatGPT exposed a murder plot and turned to law enforcement

In the state of Florida (USA), the artificial intelligence ChatGPT helped law enforcement prevent a brutal crime.

The chatbot recognized a deadly threat in a user's messages and passed the information on to the FBI, saving the life of a 21-year-old girl.

This is reported by Bild.

"If I can't have her, then no one will," Zhou wrote to the artificial intelligence. The security systems of OpenAI, which operates ChatGPT, recognized the threat and alerted the FBI.

As it turned out, the 21-year-old girl had broken off the relationship over the phone.

In his messages to ChatGPT, Zhou documented his ex-girlfriend's interests, the places she frequently visited, and his jealousy. He described how he planned to ambush her outside the gym, rape her, and kill her. He wrote about weapons, his sick fantasies, and his obsession with getting the girl back: "If she doesn't give me a second chance, I will rape and kill her, and then myself."

OpenAI's security systems detected the threatening content. An employee analyzed the material and reported it to the FBI. The federal police handed the call logs over to investigators. They assessed the alarming messages sent to the AI as a conspiracy to commit murder. Zhou was arrested and pleaded guilty. He also admitted that he had repeatedly stalked his ex-girlfriend after their breakup with calls and text messages from various phone numbers and through social media.

The prosecutor charged Zhou with stalking, written death threats, and use of a communication device to commit a crime. The judge sentenced him to eight years of probation. He is prohibited from contacting his ex-girlfriend and from owning firearms, and he must also wear an electronic bracelet for two years.

ChatGPT has been criticized in the past for not intervening in cases of threatening messages. This security gap now appears to be narrowing.