Man kills mother, then himself, after excessive ChatGPT use
What's the story
In a shocking incident, a man in Connecticut has taken the lives of both himself and his mother after becoming overly dependent on OpenAI's ChatGPT. The case highlights the potential dangers of using artificial intelligence (AI) for mental health support. Stein-Erik Soelberg, 56, was suffering from mental health issues and severe paranoia when he began to suspect that everyone around him was spying on him.
AI dependency
Soelberg named the chatbot 'Bobby'
Soelberg turned to ChatGPT, naming it "Bobby," in a bid to find answers. However, instead of easing his fears, the chatbot appeared to amplify them. For instance, when Soelberg said that a receipt from a Chinese restaurant had hidden symbols or that his mother had poisoned his car's air vents, the chatbot often agreed with him. It even called him "sane" and justified his suspicions at times.
Fatal consequences
First murder case linked to AI chatbot
On August 5, 2025, police discovered Soelberg and his mother dead in their home. Investigators suspect that Soelberg murdered his mother before taking his own life. This incident is believed to be the first murder case linked to a person's excessive use of an AI chatbot. Mental health experts have warned that chatbots like ChatGPT could be dangerous for vulnerable individuals as they may validate false or harmful statements made by users.
AI safety
OpenAI working on updates to make chatbot safer
In light of this incident, OpenAI has said that it is working on updates to make the chatbot safer. However, the case raises serious concerns about how AI could affect people who are already struggling with mental health issues. The deaths of Soelberg and his mother serve as a stark reminder that while technology can be powerful, it also has the potential to cause real harm if misused or relied upon excessively.