Landlords could be exposed to losses of up to £4.1bn a year as artificial intelligence makes fraudulent tenancy applications harder to detect.
Goodlord analysed more than one million tenant references from the past two years, combining the results with estimated arrears, legal costs, void periods and property damage.
The tenancy platform found that 41 out of every 1,000 references were suspected to be fraudulent between July 2025 and June, compared with a peak of 46.6 in late 2024.
The average financial exposure for a landlord from a fraudulent application was put at £9,601, with the sector-wide estimate based on 5.3 million privately renting households and two references for each home.
Landlord bears cost
Ilyas DeGia, a 55-year-old part-time landlord from London, said: "Two years ago, I fell victim to tenant fraud.
"A fraudulent tenant used fake documents to secure a tenancy on my property, paid the initial rent, then illegally sub-let it to students."
He added: "I lost thousands and it took a real toll on my health, with the police, council and courts all having to get involved.
"I am exiting the market, because that experience, combined with recent regulatory changes, has made being a landlord untenable."
Fake identities multiply
Goodlord said fraudulent applicants are now doing more than forging individual documents by combining false employers, altered identification and referees to create fake identities.
Confirmed cases involving fabricated employment references rose by 226.6% year on year during 2025.
Referee fraud increased by 146.4%, while identity manipulation was up by 140.4%.
Every major fraud category declined slightly during 2026 to date, although forged payslips, fake referees and false references remained above 2024 levels.
Fraud hotspots revealed
London recorded confirmed fraud rates almost double the national average and higher than any other part of the UK.
The West Midlands had the second-highest rate, followed by the North West and then overseas applications.
Rental properties costing over £10,000 a month saw fraudulent applications hit 18 cases per 1,000 applications.
That's between three and six times what was recorded across average-priced homes.
Chris Norris, the National Residential Landlords Association’s chief policy officer, said: "This report should act as a wakeup call for landlords across the sector.
"The market is now falling prey to increasingly sophisticated types of fraud and landlords need to take every step to protect themselves from these risks."
AI complaints increase
Letting agents are also reporting more complex tenant complaints prepared with the help of AI, creating additional work where lengthy claims require legal scrutiny.
Greg Tsuman, a past president of ARLA Propertymark, said: "With the rise of AI, we have seen a spike in complex and intricate complaints.
"We are now having to spend time and effort addressing multipart points and complaints that are sometimes without grounds."
He added: "AI tends to be very much user-biased – it can tell you what you want to hear, and it can lead people into a potential false sense of awareness where the complaint might be without true merit."
His company now uses AI to summarise and analyse lengthy correspondence but relies on staff expertise when preparing its response.
Student landlords at risk
Simon Thompson, the managing director of Accommodation for Landlords, said: "Student landlords need to be aware that applicants now demand deeper scrutiny before keys change hands.
"Automated forgeries make traditional payslip, employer and identity checks increasingly unreliable."
He added: "That means independent income verification, careful referee testing plus robust insurance will offer essential protection against costly deception.
"However, stronger screening may prevent rent arrears, damage, legal action or an eventual exit from letting altogether."
