To address the growing cross-channel threat, Birmingham, Ala.-based Regions Bank recently implemented a new enterprise fraud-monitoring system aimed at detecting fraud in areas such as debit cards and online banking, according to Wallach. While he declines to offer specifics about the solution, Wallach says it is connected to feeds from various channels, giving the bank an enterprisewide view of fraud and allowing it to write rules and score activities across product lines and channels.
"The entire fraud management process is centered around analytics, rules and neural scoring models," Wallach relates. "The approach there hasn't really changed. We've had that approach in individual silos for years now. It's just enriching the data in all those different aspects from various products and channels."
Phishing and key-logging attacks, for example, target multiple channels, Wallach suggests. In a typical phishing attack, for example, a hacker may be able to compromise a customer's PIN in addition to his or her online banking credentials. "[Fraudsters] are obviously going to hit on both those channels, so if you see a particular type of fraud on one channel, you can safely assume that you're going to see it on another channel," Wallach details. "You can immediately react to that in an enterprise system."
The challenge, however, is in implementing a system that is both tightly integrated with a bank's various channels and flexible enough to react to change, according to Wallach. "You can't just take a system ... out of the box, unwrap it and then expect it to meet all the expectations outlined in your business case," he contends. "You have to tightly integrate the system, and you don't want to assume that the integration is only relevant to areas where fraudsters are hitting you today. As soon as they figure out what your controls are, they are obviously going to move."
The process for identifying and eventually investing in an enterprise fraud detection system began at Regions Bank about two years ago, Wallach reports, adding that initial channels went live on the new system in June 2009. During the process, he notes, the business case for the fraud solution held strong through the economic crisis. "If the business case was not as strong as it was, the economics that we all dealt with might have hindered the situation," Wallach recalls. "In light of the business case, the changing economy and the fraud that is indicative of that change probably actually helped us to [justify the investment]."
Fortifying the Defenses
But while an enterprisewide view of transactions can help financial institutions identify fraudulent activities, banks still must drill down into specific channels and deploy specific point solutions to prevent fraud in the first place. Hackers may obtain user credentials for multiple channels, but they still must exploit those credentials on a channel-by-channel basis.
At Bank of Hawaii, key-logging and other malware attacks increasingly target the bank's commercial clients. "It's grown significantly in the last two years," the bank's Alama says. "We've noticed much more aggressive efforts to acquire credentials on the [commercial] side, whereas before it was primarily concentrated on the retail side."