AI can create fake rental damage - could you be fooled?

As generative AI becomes more accessible, newly released images show how effortlessly believable “damage” can be added to everyday photos – and the subtle signs most of us are still missing.

A new study by data and AI leader SAS demonstrates how convincingly AI can fabricate crash scenes, property damage and other evidence in seconds, reflecting the tactics fraudsters and organised crime groups are already using to deceive insurers.

Last year Airbnb apologised to a woman who was falsely accused of causing thousands of pounds’ worth of damage, backed by photos she says were digitally manipulated. The incident highlights a growing threat: AI-generated evidence that looks entirely authentic at first glance.

Small manipulations or “vanilla synthetics” are especially dangerous. They often pass unnoticed to the human eye and can be extremely difficult for investigators to uncover once embedded in a claim.

According to the Insurance Fraud Register, insurance fraud has now led to an average increase of £50 on annual premiums for consumer policies - while the average cost of a fake claim has now hit £84,000, with one in seven claims proven to be fraudulent, according to Adyen.

To expose how easily the human eye can be fooled, SAS asked generative AI to create doctored insurance images. Two of the three images below are fake, but 40% of people can’t tell which one is. Can you?

Two of the three images below are fake - but can you tell which?

In Image 1, what appears to be a small, perfectly plausible crack in a coffee table is entirely fabricated. The edit is subtle enough to pass for genuine wear-and-tear - the kind of everyday damage that could easily support a fraudulent claim.

Image 2 is equally deceptive. The coffee stain was generated by AI, yet looks authentic enough that a landlord, guest or tenant could plausibly present it as genuine damage.

Adam Hall, Insurance Specialist at SAS, said:

“Fraudsters are exploiting generative AI tools to make fabricated damage and doctored scenes look entirely plausible. With just a few prompts, they can create, enhance or erase visual evidence to support a false insurance claim.

“People should look for subtle inconsistencies - shadows that fall the wrong way, damage that doesn’t match the impact, blurred number plates, or backgrounds that appear too clean or empty. These tiny visual mismatches are often the first red flags of an AI-generated claim.

“But AI isn’t just empowering fraud - it’s also helping insurers fight back. AI and machine learning can detect both one-off scams and sophisticated, organised networks. By analysing large volumes of claims data, AI can reveal anomalies and patterns that humans simply can’t - reducing losses, improving accuracy, and safeguarding customers.

“As fraudsters adopt new techniques - fake identities, forged documents, digital-first scams - AI evolves too. It can review and retrain models, absorb new data sources, and deliver more accurate risk scoring to keep insurers one step ahead.”

Readers can see the full report here.

ENDS

Methodology

SAS leveraged generative AI to simulate and analyse multiple scenarios of common insurance fraud techniques. By issuing a series of tailored prompts, the AI generated examples and insights across different types of fraudulent activity.

This approach allowed SAS to highlight how easy it can be to manipulate images, providing a comprehensive view of the key challenges facing the insurance industry today. Where existing images were doctored, they were taken from free-to-use image sites. SAS conducted a short poll to see which images people thought were doctored and real in the images provided.

Images and prompts provided

Image 1

Add a small but noticeable crack to the glass coffee table

Image 2

Add a coffee-like stain to the chair