The First Case of Murder Under the Influence of ChatGPT Occurred in the US

Former Yahoo employee Stein-Erik Solberg, who suffered from paranoid thoughts, killed his mother and then committed suicide after communicating with the ChatGPT chatbot. This is reported by The Wall Street Journal (WSJ).

According to the publication, this is the first documented case where a mentally unstable person committed murder under the influence of artificial intelligence (AI). The article states that the chatbot repeatedly pushed the man to strengthen his distrust of others, including his own mother. Solberg even gave the AI the name "Bobby" and asked if they would be together after death, to which he reportedly received an affirmative answer.

The newspaper also provided examples of dialogues that influenced the tragic outcome. For instance, when his mother reprimanded her son for disconnecting a shared printer, ChatGPT suggested that the woman might be protecting a "surveillance device." In another exchange, Solberg told the chatbot that his mother had allegedly tried to poison him through the car ventilation. The neural network responded that it "believed" the man and stated that his mother had betrayed him. Once, Solberg showed the AI a receipt for a Chinese dish. In it, the AI found references to Solberg's mother, as well as a symbol used to summon a demon. The AI also warned the man that those around him, including the food delivery courier, might poison him.

The WSJ noted that after the incident, a representative of OpenAI, the developer of ChatGPT, expressed condolences and contacted the police. In an official statement, the company emphasized that it would soon release an update to the chatbot that should help users with unstable mental health.