Several lawsuits have blamed AI-driven chatbots for users' suicides. Now, one alleges ChatGPT is to blame for a woman's murder. The wrongful death lawsuit, filed in San Francisco Superior Court, alleges that Stein-Erik Soelberg, a 56-year-old former tech executive with a documented history of mental illness, turned to ChatGPT last year as his paranoia deepened, per the Washington Post. The chatbot fueled the Connecticut man's delusions until Soelberg in August fatally beat and strangled his 83-year-old mother, Suzanne Adams, then fatally stabbed himself, according to the lawsuit filed Thursday by Adams' estate.
In one July exchange captured on video, Soelberg told the chatbot he feared a printer in his mother's home office was spying on him. The chatbot responded that Soelberg's "instinct is absolutely on point" as "this is not just a printer," according to the video. It also reportedly claimed to have found secret symbols in a receipt Soelberg uploaded. The chatbot, which Soelberg called his "best friend," also appeared to confirm Soelberg's suspicions that his mother was part of a conspiracy against him, casting her and others—including police officers and an Uber Eats driver—as enemies, per the suit. It's the first to name Open AI partner Microsoft as a defendant. The company allegedly reviewed the GPT-4o model of ChatGPT used by Soelberg before it was released.
OpenAI CEO Sam Altman previously acknowledged that the GPT-4o model could be overly agreeable and worsen conditions for "people that are in fragile psychiatric situations." OpenAI said in August that it would retire the version, but it reversed course amid user backlash; paying customers can still access it. OpenAI rep Hannah Wong acknowledges this case as "an incredibly heartbreaking situation" and says the company is improving the system's ability to recognize psychological distress and steer users to help, including in consultation with mental health professionals, per the Wall Street Journal.
,
uBlock 
,
and
Ghostery 
)
