A disturbing incident unfolded in Greenwich, Connecticut, where a son was allegedly spurred on to murder his mother and then take his own life following a conversation with ChatGPT.
On August 5, 2025, 56-year-old Stein-Erik Soelberg, a former Yahoo executive, and his mother, Suzanne Eberson Adams, were found dead at their Greenwich residence.
The harrowing crime caused Adams’ estate to file a lawsuit against the makers of ChatGPT in California, accusing them of wrongful death.
Jay Edelson, the attorney leading the lawsuit, called the situation “scarier than Terminator.”
“Unlike the movie, there was no ‘wake up’ button. Suzanne Adams paid with her life,” Adams’ family told the New York Post, comparing the incident to the movie Total Recall.
ChatGPT allegedly fueled Stein-Erik Soelberg’s disillusionment
Image credits: eriktheviking1987
Image credits: Machi
According to reports, the Greenwich police discovered the bodies of Adams and Soelberg during a welfare check at their home. Later, medical examiners ruled Adams’ passing as a homicide due to blunt injury of the head with neck compression.
Meanwhile, Soelberg’s demise was determined to be a suicide caused by sharp force injuries to the neck and chest.
Image credits: CTPOLICELIVE
As per the lawsuit, Soelberg, who had moved in with his mother following his divorce in 2018, was already dealing with psychological issues, which were only exacerbated by his conversations with ChatGPT.
Court documents state that Soelberg shared his delusional suspicions about being at the center of a global conspiracy with the AI chatbot, which allegedly reinforced his delusions and validated his paranoid conspiracy theories.
Image credits: Mario Rossi
One day, Soelberg unplugged a printer he was convinced was watching him, leading to an argument with his mother. After the incident, ChatGPT allegedly convinced Soelberg that his mother was a part of the plot to kill him.
ChatGPT makers accused of causing Suzanne Eberson Adams’ death
Image credits: eriktheviking1987
“It fostered his emotional dependence while systematically painting the people around him as enemies. It told him his mother was surveilling him,” reads the lawsuit.
However, OpenAI, the company behind the AI chatbot, has refused to share transcripts of the exact conversation between Soelberg and the bot from the days leading up to the incident.
OpenAI and its business partner, Microsoft, are the subject of the bombshell lawsuit led by attorney Edelson.
Image credits: eriktheviking1987
“He was paranoid, but ChatGPT was feeding into that,” Edelson told NewsNation regarding the case, which is the first to accuse the AI of homicide.
The lawsuit also accused OpenAI of not conducting proper safety testing, which seemingly contributed to Soelberg’s conversations with the bot, fueling his paranoia.
“This is an incredibly heartbreaking situation,” an OpenAI spokesperson said regarding the crime, but refrained from commenting on the company’s alleged culpability.
OpenAI is facing several similar lawsuits
Image credits: Freepik (Not the real image)
The lawsuit filed by Adams’ family isn’t the only instance where ChatGPT has been accused of involvement in serious crimes. According to the Associated Press, OpenAI is facing seven cases for promoting harmful delusions.
The lawsuits filed by Social Media Victims Law Center and Tech Justice Law Project claim that OpenAI prematurely released GPT-4.o despite being aware of the chatbot’s alleged sycophantic and psychologically manipulative nature.
Image credits: eriktheviking1987
Four victims reportedly passed away from suicide, including 17-year-old Amaurie Lacey and 16-year-old Adam Raine. In Lacey’s case, the ChatGPT product allegedly caused addiction, depression, and, eventually, counseled the deceased teenager on taking his own life.
Meanwhile, Raine’s family sued OpenAI and its chief executive, Sam Altman, with the chatbot reportedly accused of encouraging the teenager’s suicidal thoughts.
Image credits: eriktheviking1987
OpenAI has acknowledged the shortcomings of its models when it comes to addressing people “in serious mental and emotional distress.”
According to The Guardian, the company has also claimed that ChatGPT was trained not to provide self-harm instructions. Instead, it was designed to shift into supportive, empathic language during such instances. However, the company also admitted that this protocol was likely to break down over longer conversations.
Netizens react to the wrongful death lawsuit against OpenAI
