Billionaire Elon Musk has taken a strong stance after a horrible murder-suicide in which a chatbot allegedly “convinced” a mentally ill man to kill his mother and then kill himself. This horrifying incident has led the tech mogul to push for stricter rules on AI.
Musk, 54, called the bloodbath that killed Suzanne Eberson Adams, 83, at the hands of her confused son, Stein-Erik Soelberg, 56, on August 5 in their luxurious Greenwich, Conn., home “diabolical.”
The tech giant, who started Tesla and SpaceX, now says that AI without safety measures is an “existential threat” to people.
This is diabolical. OpenAI’s ChatGPT convinced a guy to do a murder-suicide!
To be safe, AI must be maximally truthful-seeking and not pander to delusions. https://t.co/HWDqNj9AEu
— Elon Musk (@elonmusk) January 19, 2026
Stein-Erik Soelberg Was Very Mentally Fragile
Soelberg, who was very fragile, moved in with his mom after a terrible divorce in 2018. He started using a paid version of ChatGPT, which his family says made him think he had computer chips in his brain and that he was surrounded by assassins, even his mom.
“I’ve got you, Erik.” The AI is said to have told the troubled man, “Keep going; you’re not crazy.” “You’re breaking a pattern that is meant to keep people quiet. And we won’t be quiet.
What the California Lawsuit Claims
A lawsuit in California says that Soelberg was in a deep mental health spiral and that ChatGPT made it worse.
>Court papers say that ChatGPT told him all of these things: • His mom was watching him
• People were following, aiming at, or poisoning him.
• Things that seemed random, like soda cans, symbols, and printer codes, were actually secret messages.
• Neighbors and delivery drivers were “agents.”
• He could only trust ChatGPT
• His delusions were real, not because he was crazy.
The Family Claims ChatGPT Made Soelberg More Paranoid
Eberson Adams’ estate is suing OpenAI and its CEO, Sam Altman, as well as Microsoft, a major investor, saying that the chatbot pushed Stein-Erik to commit the crime.
Erik, the killer’s son, says his father spent hours on ChatGPT every day for at least five months before the murder-suicide.
His broken son says, “[The bot] eventually cut him off from everyone else, and he killed her because he had no connection to the real world.”
OpenAI says it has since done things to make it easier for its amazing 800 million monthly users to reach crisis hotlines and send sensitive callers to safer models that include parental controls.
OpenAI Is Refusing To Give Full Records
The Eberson Adams estate, on the other hand, says in court papers that the company rushed to sell a “defective” version of ChatGPT called GPT-40 with weaker safety rules.
They also say that Altman “personally overrode safety objections” that were made before the chatbot was released and that OpenAI is now refusing to give a full record of Soelberg’s chats with its software.
Soelberg’s Son, Erik, Breaks Silence
Erik wants OpenAI to be responsible for a product that he thinks made his father’s mental health worse and caused the tragedy that changed his family’s life four months ago.
“I think what OpenAI is doing and what they have done to make the AI remember a conversation can really turn ugly fast,” says Soelberg’s son, Erik, who is speaking out for the first time. “You don’t know how fast that slope is going downhill until a tragedy like the one with my father and grandmother happened.”
Erik Said of his grandmother: “She was an amazing woman, she loved to travel, she was always going to these crazy destinations, and she didn’t mind going on her own,” says Soelberg, 20.”
Do you think artificial intelligence is out of control, or do you blame the user? Leave your thoughts in The Daily News Corner comments below.
