How FedNow Can Spur New Anti-Fraud Strategies

The Fed’s new real-time payments platform will pose challenges for risk management. Here’s how to cope with that.

The U.S. Federal Reserve made its FedNow instant-payments service live in July. It enables a faster flow of cash for companies and individuals, improving the overall flow of money through the U.S. economy.

But fraudsters are rubbing their hands, grinning from ear to ear in anticipation.

That’s because the service is now running online, along with exponentially growing amounts of instant payments, instant paychecks, instant bill payments, and instant government payments. Banks and financial institutions realize that if they haven’t implemented robust anti-fraud strategies, they had better, and soon.

The “instant-payments era” is now official with the backing of the U.S. Federal Reserve. The expectation is that as more banks choose to use FedNow, the need for more advanced anti-fraud systems implementations is ever more dire.

While most financial institutions are investing, or have invested, in some sort of anti-fraud system, many of them can’t keep up with real-time transaction activity. Or they aren’t flexible enough to consistently stay ahead of new fraud patterns and financial channels as they emerge.

Of the 35 early FedNow adopters, two of the largest U.S. banks, JPMorgan Chase and Wells Fargo, are running the service for instant transactions. In addition, there is a myriad of service providers that support payment processing for financial institutions, as well as financial institutions serving as settlement agents and liquidity providers, in the mix.

Immense Risk

Outstanding questions regarding FedNow relate to mitigating risk from fraud and whether banks will charge for the FedNow service. For its part, the central bank anticipates that as this service evolves, becoming more sophisticated, so will the need for anti-fraud strategies.

This is especially true as the service becomes integrated into banking and credit union apps and Web sites. In fact, some Fed officials and other experts in instant payments are predicting FedNow could circumvent the need for a central bank digital currency.

According to Juniper Research, the number of instant-payment transactions will rise from 97 billion in 2022 to over 376 billion globally by 2027, which represents an astounding 289% increase. This increase over the next five years illustrates how quickly the banking transformation process will be realized.

As money moves faster, though, regulators haven’t adapted at the same speed. So technology innovation is filling the gap. One example is the development of artificial intelligence.

Partly due to the pandemic, the United States’ long-term movement toward instant payments accelerated faster than the banking systems’ ability to keep up. Fulfilling the demand from merchants and consumers for real-time transactions opened up a black hole of immense fraud risk.

Unfortunately, it has become clear that financial institutions’ capability to process vast volumes of data swiftly— without the aid of a fraud-prevention solution enabled by artificial intelligence (AI)—will be insufficient to protect either the institutions or their customers from fraud.

The Hybrid Approach

Fintech is constantly evolving as banks look to deliver digital conveniences and options for consumers. According to Juniper Research, cross-border transactions will go from 631 million payments globally in 2022 to over 6 billion in 2027. This development means the handling of rapidly inflating volumes of data is crucial.

All constituents involved, and especially those related to instant payments, require reliable fraud-prevention strategies to precisely process immense volumes of transactions in real time. The accelerated pace of processing doesn’t allow much time to scrutinize potentially fraudulent activity—something that has become one of the leading concerns of the banking community.

For some time now, bank accounts have been cybercriminals’ key targets. Using such tactics as phishing and spear phishing, attackers have become increasingly clever in obtaining vital information on their victims. They later exploit the victim’s information to gain access to their bank accounts, either directly or by tricking victims into making transfers into the attackers’ accounts willingly.

Multi-channel fraud is becoming the most favored type of fraud among criminals who divvy up their attacks across multiple channels. For example, when attacks occur across common channels like online banking, mobile transactions, Single Euro Payments Area (SEPA) transfers, and customer-service contacts, many of the fraud-detection mechanisms that don’t leverage AI are often challenged with identifying complex patterns. That’s because their focus usually is on the user behavior of a single channel, not a multitude.

These types of payment-monitoring platforms often prove to be ill-equipped to detect multi-channel fraud attacks. This is why financial institutions are turning to more robust hybrid anti-fraud systems that can use AI for faster and more thorough detection.

Mitigating fraud effectively requires the use of real-time monitoring solutions. To counteract various forms of fraud without impeding the speed of transactions—and, ideally, to halt the fraud without financial loss—most banks are now turning to hybrid AI fraud-detection solutions.

To be effective, such systems use advanced fuzzy logic and intelligent profiling for a human-like approach towards data analysis to flag suspicious activities in milliseconds. With AI integration, it is also beneficial to have capabilities to customize fraud-detection rules and run simulations to conduct extensive testing before deployment.

Start Now

To keep up with the speed of FedNow, advanced analytical and reporting dashboards are recommended to provide a quick overview and analysis. Hybrid systems continuously track user behavior in real time and analyze it for any irregularities, with machine-learning methods playing a significant role.

However, it is important to keep in mind that the effectiveness of these solutions is primarily dependent on the maturity of the machine-learning models they rely on. Traditional models typically need thorough training before they can be deployed with a reliable accuracy rate—unfortunately, time that banks don’t have.

Hybrid AI technology provides a comprehensive approach that enables a more profound and real-time analysis of transactional data, mitigating risk amid the high-speed conditions of instant-payment systems.

Instead of relying solely on data-driven machine-learning methods, hybrid AI technology blends machine learning with knowledge-driven techniques like fuzzy logic-based scorecards and watch lists, as well as dynamic profiling.

This blend allows for effective fraud detection even when data are complex or imprecise. The great advantage is that the solution can be effectively deployed practically from the very beginning.

Summing up

In summary, in this new era of banking, the financial industry must keep up the pace because:

1. Instant payments are on a rapid ascent, with global transactions expected to exceed 376 billion by 2027;
2. There are more substantial real-time fraud risks than ever before—a challenge current banking systems must solve;
3. Hybrid AI technology is one of the few solutions on the market to solve complex challenges with fraud;
4. The need to conduct profound, real-time transactional data analysis, and detect fraud, even under high-speed conditions, is critical;
5. Deployments of Hybrid AI technology doesn’t require an extensive training period, meaning it can be effectively deployed from the outset.

As the need for innovative, real-time solutions to fraud prevention becomes increasingly clear, Hybrid AI, with its powerful detection engine and comprehensive approach, is a missing piece of the puzzle for navigating FedNow’s challenges. These are exciting times, but keep in mind the importance of being fully prepared to tackle the challenges of fraud while reaping the rewards of instant payments.

—Roy Prayikulam is senior vice president for risk & fraud, RiskShield, at INFORM.