Fraudulent credit card transactions are a pressing concern among retailers, card issuers, and consumers alike. With a reported 31.8 million U.S. consumers experiencing credit card fraud in 2014 alone, sweeping changes have been put in place to thwart its devastating effects. Technology enhancements like EMV chip-enabled cards, fraud alerts from card issuers and banks, and dozens of no-cost credit monitoring services all work together to fight the ongoing issue of credit card fraud throughout the country. Despite good intentions, some consumers desire more from card issuers in accurately detecting – and halting – fraudulent activity on credit or debit cards.
The majority of consumers can share at least one awkward moment when a credit card purchase was declined in error. The cashier hands back the card with a pouty look, and in most cases, shares discretely that the transaction can’t be completed with the credit card provided. The account isn’t near its credit limit, was just used at the gas station down the road, or has ample funds available to cover the declined purchase amount, and still, individuals are forced to leave the store without their purchase in hand. All too often, the decline of a credit or debit card based on information received by the card issuer that triggers a potential fraudulent purchase isn’t accurate. Consumers are left with an incorrect, inconvenient stop placed on an account based on information that isn’t quite as intelligent as is should be.
Decision Intelligence from MasterCard
Although the technology behind artificial intelligence is hardly a new concept, it is now being used in different applications to benefit businesses and individual consumers on a higher level. MasterCard, one of the leading issuers of credit cards throughout the country, recently announced the development and implementation of an artificial intelligence tool, Decision Intelligence. The service, touted as a comprehensive decision and fraud detection program, utilizes artificial intelligence technology to allow financial institutions the ability to accurately process transactions and detect fraudulent purchases for customers. The intent of the program is to decrease the number of false positives that plague consumers when the detection of credit card fraud is less than accurate.
According to a recent study conducted by Javelin, nearly 15% of all consumers with an active credit or debit card experienced a false positive in fraud detection over the last year. While the experience is often equally embarrassing and frustrating to the buyer, the research highlights an even greater problem for consumers and credit card issuers. A number of purchases declines racked up to more than $118 billion in lost sales – but only $9 billion of those transactions were actually found to be linked to fraudulent activity. The extent to which false positives in identifying fraudulent credit and debit card activity poisons the retail system is costly from all perspectives.
To combat the issue of false positives, Decision Intelligence by Mastercard is meant to infuse the technology of artificial intelligence with already-established proprietary evaluations of credit use by card members. Through the use of sophisticated algorithms, Decision Intelligence works to provide credit card issuers a predictive score based on the consumer’s past behavior. Countless data points are used to develop analytic information about a transaction at the moment a card is swiped or tapped, and the score is generated and used to determine if a transaction is indeed fraudulent or not. Artificial intelligence technology allows MasterCard to build on previous transaction scores to help understand each individual card holder and ultimately reduce false positives.
How AI Differs from the Current Landscape
Most card issuers have a system in place to detect potentially fraudulent activity based on a card holder’s previous purchases. Not only is data used to collection information on purchase amounts, frequently visited retailers (both physical locations and online), and locations, but it is also used to analyze how credit card holders spend on a specific card over time. If a card holder makes an unusual purchase or attempts to buy an item in a store that is hundreds of miles away from home, a fraud alert may be placed on the card and the transaction is ultimately declined. The problem with this method is that card issuers are utilizing rules-based thresholds that have inherent restrictions. False positives are the all-too-common outcome.
MasterCard’s use of artificial intelligent in detecting and ultimately stopping fraudulent card activity is the first worldwide application of such technology. However, other card issuers have stepped up their game in terms of using advanced methods to deliver more accuracy in credit card fraud alerts in the past few years. Visa, for instance, has implemented its Advanced Authorization System to identify fraudulent activity on cardholder accounts. Large sets of data are evaluated systematically to determine what constitutes legitimate spending and questionable purchases from each of its card holders. Similarly, PayPal has embraced a level of artificial intelligence for its customers to detect fraud in recent years. Currently, though, MasterCard stands alone as the leader in artificial intelligence technology use for card holders on a grand scale.
One of the issues presented to smaller card issuers is the ability to fund more accurate techniques for identifying fraudulent transactions among card holders. The use of big data is critical to the process of using artificial intelligence in an accurate, meaningful way, but the widespread use of consumer data puts it just out of reach from a cost perspective for smaller companies. Businesses now use data to determine buyer preferences and tendencies both in-store and online, and they form that data into relevant marketing to pinpoint target segments of buyers. That equates to a high price tag, and less visible players in the credit card issuer marketplace have less access because of it, making it difficult to keep up with powerhouse companies with Visa, PayPal, and Mastercard.
The future of credit card practices is bright, especially from a retailer and card issuer perspective. However, taking away the annoyance of false positives when it comes to potentially fraudulent activity is a clear win for consumers, too. As more companies gain access to the data needed to run progressive artificial intelligence programs like Decision Intelligence, individual consumers, merchants, and the almighty card issuers will benefit immensely.