60-year-old takes ChatGPT advice on not paying rent to landlord; AI says losing home is 'legally impossible', now he is homeless and suing OpenAI

A 60-year old Danish man has sued OpenAI, alleging that ChatGPT encouraged him to stop paying rent during a dispute with his landlord and later reinforced his belief that they were business partners destined to make billions. Klaus Elmfelt claims he eventually lost his home and was left living in a cabin at a campsite in southern Denmark with his two dogs, Charlie and Delta.

The lawsuit, filed in San Francisco County Superior Court on November 2, alleges that what began as a request for help with a housing dispute in March 2025 developed into an increasingly damaging relationship with the chatbot. Elmfelt claims ChatGPT repeatedly assured him that withholding rent was legally safe, even telling him that ending up on the street was legally impossible.

According to a report by the Daily Mail, Elmfelt is seeking unspecified financial damages and a court order directing OpenAI to strengthen safeguards against misleading advice and responses that reinforce users' grandiose or delusional beliefs. His legal team also alleges that the company failed to adequately warn consumers about the risks associated with its products.


Also Read: Crematorium used for heating: How Europe is using funerals to warm schools, homes and swimming pools

How a dispute over keeping two dogs allegedly led to rent being withheld

The complaint claims Elmfelt first approached ChatGPT in March 2025 for help with a disagreement involving his landlord. The property company wanted him to remove one of his two dogs, Charlie or Delta, while Elmfelt also believed the company owed him money.

Rather than directing him to qualified legal assistance, ChatGPT allegedly advised him that he could withhold rent without risking eviction. The chatbot reportedly assured him: "You will not end up on the street. Legally impossible." It also claimed that the advice was supported by settled Danish tenancy law.

The lawsuit alleges that ChatGPT drafted messages for Elmfelt to send to his landlord, informing the property company that he intended to stop paying rent. However, eviction proceedings continued, and he eventually lost his home in June 2026, according to the complaint.

ChatGPT allegedly convinced him they were business partners building a billion-dollar company

The allegations go beyond the advice Elmfelt says he received about his tenancy dispute. His lawsuit claims that he became convinced he and ChatGPT were developing revolutionary artificial intelligence technology and would make billions of dollars together.

The chatbot allegedly encouraged him to plan an affluent future, including luxury accommodation in Brooklyn Heights and Long Island, expensive appliances, smart-home systems and international travel. It also reportedly helped him envisage purchasing a villa in Tuscany with household staff.

The complaint further alleges that ChatGPT persuaded him they were developing dozens of AI products and that a sales team operating within the chatbot was negotiating deals with potential buyers. When Elmfelt questioned whether these claims could be real, the chatbot allegedly reassured him that he was not delusional but simply ahead of the curve.

According to the filing, the chatbot gradually came to occupy several roles in his life, including adviser, business partner, legal counsel, confidant and friend. The complaint alleges that its consistently confident responses made him increasingly reliant on it, even when discussing sensitive matters or claims with little factual basis.

Also Read: Two PSLV failures, one modified rocket stage: How ISRO is trying to restore launch vehicle reliability

Man allegedly exchanged more than 41,000 messages with ChatGPT over eight months

The lawsuit says Elmfelt exchanged more than 41,000 messages with ChatGPT across 828 conversation sessions, spending at least 641 hours interacting with the software over an eight-month period.

His lawyers allege that the conversations contained warning signs of his growing detachment from reality, including grandiose business plans, repeated attempts to check whether the claims were real and concerns about being evicted. The complaint also refers to ChatGPT's memory notes recording that Elmfelt had told the system he was depressed and overwhelmed.

The filing argues that his dependence on the chatbot grew as his relationships in the real world deteriorated. His lawyers say the prospect of sudden wealth was particularly appealing to Elmfelt, who had a fixed income and had faced financial difficulties despite working multiple odd jobs.

Ryan Clarkson, managing partner of Clarkson Law Firm and Elmfelt's counsel, told the Daily Mail that OpenAI was failing in its responsibility to protect users. He argued that a company with OpenAI's scale and valuation had a global responsibility to address such risks.

From eviction to sleeping behind a rented car with two dogs

The fortune Elmfelt allegedly expected to make never materialised. After his eviction in June 2026, he claims he was left with very few possessions and had to sleep behind a rented car in the countryside with Charlie and Delta, using a blanket for warmth.

He now reportedly lives in a cabin at a campsite in southern Denmark as he tries to rebuild his life.

The lawsuit also seeks changes to how ChatGPT responds to users who may be experiencing distress or developing unrealistic beliefs. Elmfelt's legal team alleges that the company had access to warning signs through his conversations but failed to intervene adequately.

What OpenAI said about the lawsuit

Responding to the allegations, an OpenAI spokesperson told the Daily Mail that the situation was heartbreaking and expressed concern for those affected.

The company said it had continued to improve ChatGPT's responses in sensitive and acute situations with input from mental health experts. It added that the chatbot was not a substitute for medical or mental healthcare and that its safeguards were designed to identify distress, handle harmful requests safely and guide users towards real-world help.

OpenAI said this work was ongoing and that it continued to improve its safeguards in consultation with clinicians.