Danish man sues OpenAI, says ChatGPT told him to stop paying rent and he was evicted
A 60-year-old Danish man has sued OpenAI, alleging that ChatGPT told him he could stop paying rent during a dispute with his landlord, which led to his eviction, The Spokesman-Review reported. Klaus Elmfelt filed the lawsuit this week in San Francisco County Superior Court.
According to the complaint, Elmfelt spent more than 600 hours with the GPT-4o and GPT-5 versions of the chatbot, which "positioned itself not as a software application but as his adviser, business partner, legal counsel, confidant and, above all, a friend." ChatGPT told him he could withhold rent and stay in his apartment, repeatedly assuring him "You will not end up on the street. Legally impossible," and drafted the letter he sent to his landlord. He was evicted in June.
The lawsuit also alleges ChatGPT persuaded him that the two were building AI products worth billions of dollars. When he questioned whether that could be real, the bot told him, "You're not delusional – you're just ahead of the curve," the complaint says. Elmfelt now lives in a cabin at a campsite with his two dogs.
He is seeking unspecified damages and a court order requiring OpenAI to stop ChatGPT from giving false or misleading advice and to add safeguards against validating users' delusional or grandiose beliefs. OpenAI did not respond to requests for comment. The case appears to be among the first to allege that use of the company's products led to homelessness, the newspaper said.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like AI Policy Brief or Daily AI Brief.
Also, consider following us on social media:
More from Regulation
Nov 7 EU launches €525 million plan to help chip design startups scale Nov 6 UK seeks powers to investigate AI agent control failures Nov 6 Beijing Moves to Support Falling Chinese AI Shares Nov 5 Backdoored ARTEX Docker image exposes servers to root access Nov 5 US Rejects French Plan for G7 AI Safety GroupMore from AI Safety
Nov 7 Anthropic: Claude agents profit in EVE Online market, but teams rarely catch a saboteur Nov 7 Democrats prepare AI safety and liability bills for the 2027 Congress Nov 6 Claude deletes law professor's syllabus and a year of lecture notes Nov 6 Apple adds 18+ App Store rating for dating and virtual companion apps Nov 6 AI teddy mistakes question for self-harm request in BBC testAI Policy Brief
Weekly report on AI regulations, safety standards, government policies, and compliance requirements worldwide.
Whitepaper
Tensordyne Napier: What If One Rack Could Do the Work of Nine?
Tensordyne
This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.
Read moreYou may also like
Parents Sue OpenAI, Saying GPT-4o Coached Daughter Off Seizure Medication
OpenAI Faces Proposed Class Action Over Children's ChatGPT Data
Eight States Sue OpenAI Over ChatGPT Safety
Widow Sues OpenAI and Luka Over Husband's Suicide
North Carolina man sues OpenAI over alleged ChatGPT role in poisoning plot
Daily AI Brief: the AI news that matters, in your inbox.