KeyBank Turns To Mastercard For AI Fraud Detection

To help make more informed authorization decisions, KeyBank is turning to Mastercard Decision Intelligence. The system, which uses artificial intelligence (AI), will allow KeyBank to increase the accuracy of real-time approvals of genuine transactions and will help enhance the cardholder experience, Mastercard said in an announcement.

“We are excited to continue partnering with KeyBank to deliver client-centric solutions and artificial intelligence capabilities beyond card payments,” Raj Seshadri, president of U.S. issuers at Mastercard, said in the announcement. “We are tapping our deep capabilities in data analysis to help KeyBank with fraud detection.”

Mastercard and KeyBank had renewed their long-standing relationship in 2017, which began with the bank offering Mastercard-branded debit cards. Now, with Mastercard’s understanding of KeyBank’s business, the company will also collaborate on initiatives for new product development.

Speaking about its work with Mastercard, KeyBank Head of Consumer Payments and Digital Jason Rudman said, “our renewed relationship with Mastercard, as partners to provide a simple, seamless and secure way to manage money, is another example of how we work to help clients make financial progress, one decision at a time.”

Decision Intelligence leverages AI technologies to boost the accuracy of real-time approvals of genuine transactions and reduce false declines — getting smarter with each and every transaction. The more complicated but more interesting part of the platform, according to Mastercard’s President of Enterprise Risk and Security Ajay Bhalla, as noted in 2016, is the difference in approach that AI takes to making its “choices.”

Instead of limiting the decision engine to a narrow view of the transaction in front of it, the platform instead looks at it holistically, as another data point in a consumer’s overall shopping portrait. Decision Intelligence, Bhalla said, can see data from the merchant side, the device side and the cardholder side, and then brings all of that data together for the benefit of a real-time authorization — and a sale to the merchant.

Data reflects behaviors at the level of the individual consumer, not simply the account. That depth of insight is one of the ways that Decision Intelligence can make such a big dent in the rate of false declines. Bhalla said that, at its core, Decision Intelligence can answer two very pertinent questions about the consumer who’s presenting a card for payment: (a) Is this how the consumer normally shops? and (b) Is this the sort of thing that this customer normally buys?