Payment Rails Challenge: Faster Payments, Faster Fraud
- Payment rails are streamlining money transfers from payer to payee
- The Federal Reserve has introduced the FedNow service
- Faster payments are great, but also mean a smaller window of opportunity to detect fraud
New and innovative payment rails -- the payment platforms or a payment network that moves money from a payer to a payee -- are designed to bring some of the biggest legacy pain points of the financial services industry down to size. Fast, smooth passage of funds is a plus for all sides of a transaction.
As reported at PYMTS.COM, the Federal Reserve's ongoing development of a new real-time payment rail, the FedNow service, continues as it adds FinTech Finastra to its pilot program.
Federal Reserve graphic
Finastra announced this week that it will lend support to the development, testing, and adoption of the service. They combine with other banks, credit unions (CUs), and payment technology firms in an initiative to launch the first new payment rail in decades for the U.S. As PYMTS relates, Finastra's own technology helps financial institutions strengthen their own back-office operations and modernize payment capabilities, up to and including looping those entities into real-time payment schemes.
The Fraud Challenge
When it comes to addressing fraud in this space, Featurespace Senior Vice President of Global Product Management Dena Hamilton referred to lessons learned from traditional wire transfer fraud to identify the biggest threats for businesses within the payment rail environment in terms of high-dollar losses and reputational damage.
"With the introduction of faster payments, we've seen an abbreviated window in which customers are able to dispute or retract a payment," she said. That makes proactive monitoring and detection for fraud and financial crime — rather than a reactive approach — critical.
Hamilton points to the experiences of the United Kingdom and Europe as a reference:
“For our counterparts in the U.K. and across Europe as well, what we’ve seen with the faster payments initiative is that it’s reduced the recourse window as a financial institution from days to two hours,” Hamilton said. “So what that means to us here in the United States, as we move to real-time payments, is that we need to look at how we’re going to address the shorter timeframe in which we can actually identify those items that may be potentially fraudulent or suspect.”
It is truly a challenge for the payments industry to address fraud at greater speeds -- dare we say in real-time? To accomplish this, there is only one realistic type of technology that can perform the analysis of the payments and detect fraud in these short windows.
As featured in #OrboZone, processing vast quantities of data at real-time speeds make Artificial Intelligence and Deep Learning technologies absolutely essential. Powered by GPUs, these technologies are able to analyze hundreds of thousands of transactions in mere minutes, compared to CPUs which could take hours and even an entire day. These are the technologies that are being utilized by OrboGraph today in our OrbNet Forensic AI for check fraud to analyze hundred of thousands of check images to effectively detect counterfeits, forgeries, and alterations.
