In 2004, Verifeyed brought together a group of scientists and digital security experts to develop a toolkit to facilitate instant detection of digital image tampering. After seven years of research and effort, we finally created the software to accomplish this accurately and called this software Verifeyed.
Our excitement led us to immediately undertake projects for several large clients: a bank, a media publisher, and two insurance companies. Our aim was to detect fraudulent images and scanned documents to save our clients significant portions of their budgets. Having the most cutting-edge image forensics technology available and being aware of growing trends in digital frauds, we were confident about success.
After six months of effort during which we analyzed more than 1.5 million digital images, an evaluation meeting with clients revealed that most of the identified image manipulations were false positives (i.e. not real frauds), and the amount of saved money for our clients was negligible. The prevalence of digital image fraud was clear to everyone and we had the best technology, but no exciting results. We couldn’t understand what had gone wrong.
Eventually we discovered one critical point we’ missed. Despite the technology being 100% accurate in lab-controlled testing, we’d failed to account for the real-life behavior of all variables: photos, cameras with software embedded within them, and even the criminals making the frauds. Thus, we undertook a two-year process to adapt the technology to real-life conditions. We continuously communicated and exchanged data with businesses and expert representatives from different fields, bringing their feedback, evaluation results, etc. into Verifeyed. This became the most difficult part of the whole development process, but was worth it in the end.
When we went back to market after two years, the program immediately proved successful. For example, in a global insurance company we used the updated program to automatically analyze 10,000 auto claims and uncovered (verified by that insurance company) that every 1000th image associated with claims was fraudulent.
Verifeyed is a result of eight years’-worth of effort from scientists and business professionals. It makes it possible for you to use state-of-the-art technology to instantly analyze digital images with a single click. We are proud that it is used by leading intelligence and forensics services, business-oriented companies, and has received several awards from recognized business and research organizations.
Verifeyed has been awarded by the City of New York. From left to right: Babak Mahdian (founder of Verifeyed), Michael Bloomberg (Mayor of New York City), Radim Nedbal (CTO of Verifeyed) .
It is the most amazing software I’ve ever seen. It enables us to tackle digital fraud and discourage potential digital fraud from being committed. This software is bringing enormous added value to our claims solutions. It showed positive ROI almost immediately after we started using it.
Jan Stok, MBA. CEO of MAI Insurance Brokers.
In around one in 75 insurance claims, photos documenting property damage have been fraudulently retouched, says Eugene Nealon of Nealon Affinity Partners, a company based in London that advises insurers. Mr Nealon’s firm recommends Verifeyed to spot fakery.
Eugene Nealon, CEO of Nealon Affinity Partners.