MasterCard, in its first major deployment of artificial intelligence technology, is rolling out a new decision-scoring and fraud detection service called Decision Intelligence designed to help increase the accuracy of real-time card approvals at checkout, while reducing the likelihood of transactions being falsely declined.
MasterCard said the service relies on proprietary algorithms to provide a predictive score to the card issuer, based on intelligent analysis of how a specific account is used over time to detect normal and abnormal shopping spending behaviors. It also leverages account information like customer value segmentation, risk profiling, location, merchant, device data, time of day and type of purchase made.
Card issuers can incorporate this information into their existing fraud mitigation efforts, or activate a holistic MasterCard fraud-fighting tool that makes data-driven, real-time decisions tailored to the account, including defined alert and decline thresholds.
Javelin Strategy & Research estimates that in the U.S. alone, the value of false declines is more than 13 times the total amount lost to actual card fraud. MasterCard notes that more than $118 billion in sales were lost in the U.S. to false declines in 2014, and if you're a retailer or a card issuer, there must be no worse way of losing a customer than having them fall victim to a false decline of a transaction: I'm not sure what response most people would have to being falsely declined, but I know if it happened to me, I wouldn't returning to that retailer or using that card again anytime soon.
Yet while Javelin and MasterCard both note how huge and widespread a problem this has been, the risk assessment tools used to battle it seem woefully inadequate. That is particularly the case when the sector has technology at its disposal that could be used to analyze and weigh many distinct factors involved in any single transaction.
Fraud prevention seems an area ripe for AI to improve on current practices, though surprisingly, we have not heard much from the major card companies until now, even as AI quickly has become a widely studied technology (and somewhat widely deployed, though it's still early) in retail and other industries in which card payments also play an important role. Payments firm Stripe notably launched an AI-based fraud detection service recently, though it didn't specifically mention reduction of false declines as one of its aims.
The card companies certainly have brought new technology to bear in the fight against fraud, most obviously in the form of EMV chip cards, but the apparent immensity of the false decline problem should be a reminder to them that their fraud strategies need to be comprehensive and multi-layered — one advancement doesn't solve all the problems.
Overall, AI not only can improve on current fraud analysis processes, but also evolve over time by using its own machine learning capabilities to continue refining how it views and handles ongoing transactions involving the same customers and payment cards. Ultimately, it can be another tool for improving the customer experience — something it's best to work on before another false decline drives them away.