Several insurance magazines published an article last week explaining how an insured of Aegon was caught in the act of fraud. The insurer discovered the fraud by studying the photos submitted by the insured. Conclusion: the photos were taken after the date of the burglary. The photos were only analysed, however, after the insurer received an anonymous tip. The insurer was about to pay the claim.
Images are often requested by insurers, but checking whether a photo corresponds to the damage hardly ever takes place. Fraudsters submit images to substantiate a fraudulent claim knowing that the images are unlikely to be checked properly.
This form of fraud is easy to intercept! Smart software can be used to automate the screening of photos. In this way, our software automatically carries out the following checks:
Timeline analysis ⏱️
When was the photo taken? The photo was taken at an earlier or later time than that would be logically consistent with the claim. For example: a photo of a stolen laptop was taken after the burglary occurred. The software checks when the photo was taken and compares it with the date of the loss. If this does not match, then an alert is generated.
Location analysis 📍
Where was the photo taken? The photo was taken at a different location than the alleged location of the loss. For example: the photo was not taken at the insured’s address, but at the home of someone else. If the software detects that a photo location does not match what is stated in the claim, an alert is generated.
Editing analysis 👨💻
Has the photo been edited? Some fraudsters are handy with Photoshop and can adjust the amount or bank account number on a photo of a receipt or invoice. FRISS software checks whether a submitted photo is original, or whether pixels have been adjusted.
Internet search 🌐
Was the photo downloaded from the internet? For example: a photograph of a stolen bicycle is submitted, but is seems that this bicycle was never owned by the insured. An alert is automatically generated when the software can match a photo that has been uploaded with an image from the internet.
ImagePool search 🗄️
Has the same photo been submitted before? The same photo was previously submitted by another person for a completely different claim. For example: a neighbour previously submitted a travel loss claim and substantiated it with photos. The insured then does the same with his/her insurance and uses the same photo as the neighbour, but without having travelled anywhere.
Licence plate recognition 🚘
If there is a licence plate on the image, the software can automatically recognize it and then search through relevant websites. For example: a day before the damage, the car was still for sale on numerous sales sites such as Autotrader.com, Carsdirect.com, Ebay.com, etc.
We have integrated all the above techniques into our software module FRISS ImageScreening, which is designed exclusively for insurers.
Are you interested in this technique?
We are happy to discuss how we can help you use FRISS ImageScreening in your organisation’s claim process.