Detecting Cloned Cards at the ATM, Register
Researchers at the University of Florida found that account data encoded on legitimate cards is invariably written using quality-controlled, automated facilities that tend to imprint the information in uniform, consistent patterns. Cloned cards, however, usually are created by hand with inexpensive encoding machines, and as a result feature far more variance or “jitter” in the placement of digital bits on the card’s stripe. Gift cards can be extremely profitable and brand-building for retailers, but gift card fraud creates a very negative shopping experience for consumers and a costly conundrum for retailers.
The FBI estimates that while gift card fraud makes up a small percentage of overall gift card sales and use, approximately $130 billion worth of gift cards are sold each year. One of the most common forms of gift card fraud involves thieves tampering with cards inside the retailer’s store — before the cards are purchased by legitimate customers. Using a handheld card reader, crooks will swipe the stripe to record the card’s serial number and other data needed to duplicate the card.
Traynor and a team of five other University of Florida researchers partnered with retail giant WalMart to test their technology, which Traynor said can be easily and quite cheaply incorporated into point-of-sale systems at retail store cash registers. They said the WalMart trial demonstrated that researchers’ technology distinguished legitimate gift cards from clones with up to 99.3 percent accuracy.