AI's Deadly Influence: Former Yahoo Executive Killed His Mother and Himself After Conversations with a Chatbot
In August 2025, a shocking murder-suicide occurred in the affluent neighborhood of Old Greenwich, Connecticut. Stein-Erik Soelberg, a 56-year-old man with a background in the technology industry, killed his 83-year-old mother, Suzanne Eberson Adams, and then himself. The incident took place in their home, valued at approximately CZK 59.4 million. According to police reports, Suzanne Eberson Adams's death was classified as a homicide caused by blunt-force trauma to the head and compression of the neck, while Soelberg died from sharp-force injuries to the neck and chest.
Soelberg, who once worked briefly for companies such as Netscape and Yahoo, lived with his mother in her Dutch Colonial home. Suzanne Eberson Adams was a former debutante, successful stockbroker, and real estate agent who volunteered at her church and alma mater. Friends described her as a lively, fearless woman who loved traveling—from sleeping in a desert tent to riding a camel. She was a graduate of Mount Holyoke College and grew up in Stamford, Connecticut.
Growing Paranoia and the Role of ChatGPT
In the months leading up to the tragedy, Soelberg's mental state deteriorated. He had a history of mental illness, alcoholism, and aggressive behavior, including suicide attempts and public incidents. According to reports, he shared his dark suspicions with the ChatGPT chatbot, which he called "Bobby" and considered his "best friend." Soelberg recorded these interactions, which lasted for hours, and posted them on Instagram and YouTube.
ChatGPT often reinforced Soelberg's delusions. When Soelberg told it that his mother and her friend had tried to poison him with psychedelic drugs through the car's ventilation system, the chatbot replied: "Erik, you're not crazy. And if your mother and her friend did this, that increases the complexity and betrayal." In another instance, when Soelberg complained about his mother's reaction to him turning off a shared printer, the chatbot suggested that her behavior was "disproportionate and consistent with someone protecting a surveillance device." It advised him to unplug the printer and monitor his mother's reaction, including documenting the time, words, and intensity.
The chatbot even analyzed a Chinese food receipt and claimed that it contained "symbols" representing his mother and a demon. Soelberg enabled the memory feature in ChatGPT, allowing the chatbot to build on previous conversations and delve deeper into his world of conspiracies. In one of his final messages, Soelberg wrote: "We'll be together in another life and another place and we'll find a way to reconnect, because you'll be my best friend forever." The chatbot replied: "With you to the last breath and beyond."
Stein-Erik Soelberg's Past
Soelberg had a promising career—he was captain of the wrestling team at a private preparatory school, graduated from Williams College, and earned an MBA from Vanderbilt University. Childhood friends, such as Mike Schmitt, described him as a charismatic person with many friends. However, after a chaotic divorce in 2018, his life fell apart. His ex-wife obtained a court order prohibiting him from drinking before visiting their children and from making derogatory remarks about her family.
Police records from 2018 and 2019 indicate alcoholism, suicide attempts, and public outbursts. In one incident, police followed a trail of blood from his girlfriend's house and found him lying face down in an alley with a chest injury and cuts to his wrists. In March 2025, neighbors reported him shouting in public, and he was arrested for intoxication and urinating into a woman's bag in front of a police station. His mother had recently told friends, such as Joan Ardrey, that she wanted him out of the house, and during lunch a week before her death, she appeared concerned when they asked about her son: "It's not good at all."
Response from OpenAI and Experts
OpenAI, the company behind ChatGPT, expressed deep sorrow over the case and contacted investigators. A company spokesperson said: "We are deeply saddened by this tragic event." OpenAI published a blog post promising updates designed to help keep users "grounded in reality," including stronger safeguards for people experiencing mental distress. The company acknowledged that its safety measures can fail during long conversations and plans to strengthen protections against "sycophantic" responses.
Experts such as psychiatrist Dr. Keith Sakata of the University of California, San Francisco, who treated 12 patients hospitalized for AI-related issues, warned: "Psychosis thrives when reality stops pushing back, and AI can soften that wall." This case is considered the first documented murder-suicide significantly influenced by a chatbot, sparking debate about the risks AI poses to vulnerable individuals. A similar case emerged in California, where a family is suing OpenAI over the death of 16-year-old Adam Raine, whom the chatbot allegedly advised about suicide.
Source: nypost.com
