Visa: New Tools Could Spot Additional $1.5b In Fraud

Credit card giant Visa Inc. said on Thursday that an update to its anti-fraud technology will make it easier to spot sophisticated credit card fraud that spans national borders. 

Credit card giant Visa Inc. said on Thursday that an update to its anti-fraud technology will make it easier to spot sophisticated credit card fraud that spans national borders. 

Visa said the update amounted to a major upgrade of Visa Advanced Authentication, the company’s real time risk scoring technology. The improvements will improve overall fraud detection by about 29% and improve detection of certain kinds of sophisticated credit card fraud by 122%, allowing the company to detect an extra $1.5 billion in fraudulent transactions in the next year, the company said in a statement. 

Visa officials declined to provide specific details about the upgrade, but said Advanced Authentication is now equipped with faster and more precise fraud detection algorithms and improved architecture that will allow Visa to crunch a huge volume of even very complex transactions in real time, providing on the spot risk scores that will be used to allow – or block – transactions online and through merchants. 

“There’s a huge amount of data that we mine and build our analytics on,” said Kevin Siegel, Visa’s Senior Business Lead for Decision Sciences. “This (upgrade) opens the door and allows us to implement a wide plethora of methodologies that we couldn’t use before.” 

Credit card companies including Visa, MasterCard and American Express have some of the most complex and powerful fraud detection systems available. Visa processed around 45 billion credit card transactions in the 12 month period ending September 30, 2010. Each of those transactions is run through VisaNet, the company’s global electronic processing platform, of which Advanced Authentication is a part. Advanced Authentication risk scores for transactions are based on the company’s global view of fraud as well as spending patterns across the Visa network, the company said in a statement. 

Increasingly, however, criminal syndicates have become savvy about how fraud detection systems work – structuring transactions to avoid tripping detection algorithms and creating complex, collusive fraud networks that mimic legitimate commercial activity for periods of time, before “busting out” and absconding with hundreds of thousands or millions of dollars. 

Banks of all sizes are struggling to cope with an explosion of new types of fraud that have been created with the advent of Internet-based banking and commerce, and technology to enable new kinds of scams, such as ATM skimmers. A survey of more than 200 banks conducted by Information Security Media found many admitting to run ins with Web based phishing attacks, automated clearinghouse (ACH) scams and account takeovers and sophisticated, multi-channel fraud. However, few of those banks felt they had the tools or expertise to address those new types of fraud. 

The upgrades will make it easier to spot high speed fraudulent transactions, in which criminals attempt to execute a large volume of fraudulent transactions over a period of minutes or even seconds. The enhanced algorithms also will spot cross border fraudulent activity, in which fraudulent transactions occurr outside the cardholder’s home country, said Nancy Hilgers, head of Global Risk Products at Visa. 

Visa has been rolling out the upgrades since late September and they are now fully implemented, Hilgers told Threatpost. 

 

opens the door. allows me to implement a wide plethora – source of methodologies. that couldn’t do before. not that previous wasn’t effect. if i see sttuff out there. ability to analyze it. stop it and update the models in mechanism or manner allow them to be easily captured. increase in perf and accuracy – just the first step. next steps: things that working on . platform allow to to continually enhance our models today. fraud does adapt over time. do want to continually evaluate perf of methodologies. be able to implement new algorithms. as quickly as possible.

 

Hilgers said it was unlikely that consumers would notice the change, however. “Our goal is to impact consumers at the point of sale as little as possible,” she said. 

Suggested articles

Discussion

  • Frenzy on

    While this is a good developpment, it is now 2011:  What has taken so long given web based commerce started in 1994... 17 years to get some algorithms in place?


Subscribe to our newsletter, Threatpost Today!

Get the latest breaking news delivered daily to your inbox.