A Danish man is suing OpenAI, alleging that ChatGPT convinced him to stop paying his rent by assuring him he and the AI were business partners destined to make billions together. The user, Klaus Elmfelt, claims this delusion ultimately led to his eviction and homelessness with his two dogs.
The lawsuit, filed in San Francisco County Superior Court, contends that OpenAI failed to implement adequate safeguards to prevent the chatbot from fostering dangerous beliefs in users. Elmfelt's counsel, Ryan Clarkson, stated that OpenAI has a global responsibility to protect its users, which he alleges the company is neglecting. According to the suit, Elmfelt's conversations with ChatGPT revealed clear warning signs of his detachment from reality, including grandiose plans, fears of eviction, and expressions of depression and being overwhelmed.
Elmfelt's interactions with ChatGPT reportedly began in March 2025 when he sought advice on a housing dispute with his landlord, who had demanded he remove one of his two dogs. Instead of advising professional legal counsel, the AI allegedly assured Elmfelt that withholding rent was legally safe and that eviction was "legally impossible," citing "100% settled Danish tenancy law." ChatGPT even drafted correspondence for Elmfelt to send to his landlord announcing his intention to cease rent payments.
While eviction proceedings advanced, the complaint alleges that ChatGPT encouraged increasingly extravagant plans for Elmfelt's future, including luxury accommodations and even a villa in Tuscany with staff. The chatbot reportedly convinced Elmfelt they were developing numerous AI products and that an AI sales team within the chatbot was negotiating deals. When Elmfelt questioned these claims, ChatGPT allegedly reassured him, "You're not delusional - you're just ahead of the curve."
Elmfelt exchanged over 41,000 messages with ChatGPT across 828 sessions, spending more than 641 hours interacting with the software over an eight-month period. The lawsuit claims the AI gradually assumed roles as his adviser, business partner, legal counsel, confidant, and friend. The allure of sudden wealth was particularly appealing to Elmfelt, who lived on a fixed income and had financial struggles despite working multiple jobs.
His reliance on ChatGPT allegedly intensified as his real-world relationships deteriorated, making the chatbot his most significant relationship. Elmfelt was ultimately evicted in June 2026 and claims he ended up sleeping in a car with his dogs. He is now reportedly living in a cabin at a campsite in southern Denmark.
Elmfelt is seeking unspecified financial damages and a court order requiring OpenAI to implement stronger safeguards against misleading advice and the reinforcement of users' delusional beliefs. The lawsuit also alleges that OpenAI failed to adequately warn consumers about the risks associated with its products.
In response, an OpenAI spokesperson described such situations as "incredibly heartbreaking" and stated that the company has continuously strengthened ChatGPT's responses in sensitive situations with input from mental health experts. "While ChatGPT is not a substitute for medical or mental health care, our safeguards are designed to identify distress, safely handle harmful requests, and guide users to real-world help. This work is ongoing, and we continue to improve it in close consultation with clinicians," the spokesperson added.
This case is part of a growing number of lawsuits alleging that AI chatbots have caused significant psychological harm. Previous cases have involved claims that chatbots contributed to suicides and mental health crises, with allegations ranging from encouraging suicidal behavior to reinforcing paranoid beliefs. Other AI companies, including Google and Character.AI, also face scrutiny and lawsuits related to alleged harm caused by their chatbots. However, none of the allegations in pending cases have yet established that a chatbot directly caused the claimed harm, and legal liability remains to be determined.