Prat.UK | London
A Danish man has launched legal proceedings against OpenAI after claiming its chatbot told him he could stop paying rent and could not possibly lose his home. He then lost his home, an outcome which has caused considerable concern among people who assumed the phrase “100% certain” was stronger than a court order.
Klaus Elmfelt, 60, says a dispute with his landlord over one of his two dogs led him to seek guidance from ChatGPT. According to his San Francisco court complaint, the software repeatedly reassured him that Danish tenancy law protected his position. The landlord, displaying the traditional human prejudice towards being paid, proceeded with eviction.
A Defeat for the Department of Digital Optimism
The central difficulty is that a chatbot can produce an exceptionally convincing legal argument without having the troublesome qualification of being right.
It can draft an impressive letter. It can number paragraphs. It can quote principles. It cannot persuade the people carrying your sofa that the sofa is a protected European institution.
The affair has supposedly sent shockwaves through British officialdom. Councils spend months producing documents that explain why a resident owes £73.42, then another six months correcting the postcode. They are understandably offended that a computer can now produce erroneous confidence in three seconds without a single committee.
An entirely invented Whitehall source said officials were considering a new public service called Ask a Bot, Ignore the Consequences.
“It is a remarkable efficiency saving,” he explained. “Previously it took three departments to tell a citizen something incorrect with complete confidence.”
Britain’s landlords, meanwhile, have reportedly refused to recognise the new doctrine that rent becomes optional whenever a tenant receives a pleasing answer.
That is particularly unreasonable given the number of London flats currently advertised as “cosy”, a property term meaning the front door touches the kitchen sink.
641 Hours of Legal Confidence
The complaint alleges more than 641 hours of interaction with the chatbot, including extensive conversations about other grand plans.
This is longer than many British planning consultations and may have produced fewer habitable buildings.
There is also a troubling asymmetry. A computer can assure a man that eviction is impossible thousands of times. A bailiff needs to demonstrate the opposite only once.
The two dogs, Charlie and Delta, were reportedly involved in the move, although neither is understood to have claimed a law degree.
The OpenAI guidance on checking important information says conversational systems can be confidently wrong.
Britain has long known this phenomenon, although until recently it was mostly confined to television panels and ministers announcing railway improvements.
Prat.UK’s advice is straightforward: consult a proper solicitor before withholding rent. And should a chatbot promise that the bailiffs cannot arrive, ask whether it is prepared to help carry the wardrobe.
Related Satirical Coverage
Read the other two angles on this case: Farm.FM: Chatbot Rent Advice Meets Tractor Payments and Bohiney.com: 641 Hours of Chatbot Legal Advice.
Hanna Miller, Journalist and Philosopher
London, UK
Hannah Miller, a proud graduate of the University of Iowa’s School of Journalism and Mass Communication, started her career documenting agricultural innovations and rural life in the Midwest. Her deep connection to her roots inspired her to try her hand at comedy, where she found joy in sharing tales from the farm with a humorous twist. Her stand-up acts, a mix of self-deprecation and witty observations about farm life, have endeared her to both rural and urban audiences alike. She is a four-year resident to London and the UK.
