Skip to main content

How Banks Can Use AI to Manage Instant Payment Fraud Risk

Graphic representing how banks can use AI to manage instant payment fraud risk and block fraudulent transactions in real time

Why Are Instant Payments a Challenge for Fraud Teams?

Instant payments, also referred to as real-time payments (RTP), aren’t just on the horizon, they’re already here. The EU agreed on new rules requiring banks to offer instant payments to their customers. The US launched the FedNow Service, which provides additional infrastructure for banks to process instant payments for their customers. At the time of this writing, more than 300 banks and credit unions have adopted FedNow. The speed of instant payment adoption is increasing as well. By 2027, real-time payments are expected to make up 27.8% of all electronic payments globally, compared to 18% last year.

While the advent of instant payments promises benefits for banks and their customers alike, it also brings the threat of increased fraud, money laundering, and terrorist financing. Faster payments mean faster fraud, so banks will need the capability to monitor payments for suspicious activity in real time. Without this capability, the effects on customers and the bank’s bottom line could be dire. After all, instant payment money is gone in ten seconds.

What Do Banks Need to Do to Protect Against Instant Payment Fraud?

The arrival of instant payments, and the new rules surrounding it, means that banks will need stronger fraud prevention capabilities, real-time fraud prevention capabilities, integrated financial crime solutions, and increased scrutiny on instant payment counterparties.

Stronger Fraud Prevention Capabilities

The new EU rules will require instant payment providers to verify that the beneficiary’s IBAN and name match, and to alert the payer to possible mistakes or fraud before a transaction is made. The new regulations will place the financial burden of fraud onto the bank, even if the payee is not a customer. This change will mean an increased risk of fraud losses for banks. To mitigate the risk, banks will need to strengthen their fraud prevention capabilities with improved systems, processes, and technology.

Real-Time Fraud Prevention

Not only will banks need to strengthen their ability to prevent fraud, but they will also need to do it in real time. The EU instant payment rule gives banks ten seconds to credit a payee's account. In that amount of time, the bank will have to screen the payment for counterparty and other risk factors, including money laundering and terrorist financing. If the bank cannot do this as fraudulent payments occur, it will be liable for the resulting fraud losses.

Integrated Financial Crime Solutions

Regulation aiming to combat fraud and related money mule activity is ramping up. To incentivize financial institutions to take the fight seriously, the UK has already introduced regulation that holds both sending and receiving financial institutions liable for the damage caused by fraud. Integrated anti-money laundering (AML) and anti-fraud solutions can help prevent money mule transactions, provide evidence for suspicious activities, and identify the connection between the predicate crime of fraud and the associated money mule activity.

Increased Counterparty Scrutiny

Banks will need to pay especially close attention to risky payment counterparties, since instant payment fraud is often a variation of typical money mule activity. In this instance, the “money mule” does not voluntarily participate in the money laundering activity, but instead becomes an additional victim of the criminals. The victim makes a legal transfer to another bank account; the criminal activity is the defrauding and deception of the victim initiating the transaction.

“While the victim plays a necessary part in completing the fraud, the transaction itself is not criminal,” said Maximilian Riege, Chief Risk Officer at Hawk AI. “This aspect is the major differentiator to money laundering activities, including those where a money mule is a voluntary player in the layering activity necessary for laundering illicit funds. This difference makes it even more difficult for the bank to detect and stop the fraudulent transaction.”

How Can AI and Other Technologies Help Prevent Instant Payment Fraud?

Regardless of the exact methodologies employed, speed, precision, scalability, pattern and anomaly detection, and false positive reduction will play essential roles in finding and stopping instant payment fraud.

Speed & Precision

With ten seconds or less to process an instant payment, banks must employ systems and technology that can perform precise checks on the payments as soon as they occur. As false payment blocking negatively affects the customer, high-precision detection is paramount. Without technology that can precisely process and analyze data at speed, banks will be left in the dust by fraudsters.

Scalability

Offering instant payments will likely increase a bank’s transaction volume and cause more seasonal and event-based payment surges. Banks need the capacity to screen and verify the payments as the volumes grow and fluctuate. Without systems and technology designed for scalability, the risk to a bank’s bottom line posed by instant payment fraud will increase.

Pattern & Anomaly Detection 

AI is effective at predicting know patterns of fraud with high precision. However, emerging fraud patterns are difficult to predict. Since it’s nearly impossible for a bank to hire enough human investigators to review all instant payments, it will need AI models built on sound data science that can detect “unknown unknowns” with a high degree of certainty. These models detect how a specific transaction deviates from billions of other transactions analyzed, as well as other transactions in a comparable context. Banks need to find and block fraud quickly and at massive scale. The best way to do this is with AI-powered anomaly detection.

False Positive Reduction

In addition to detecting and blocking fraudulent transactions, banks will need to reduce the number of false positives blocked by their fraud systems, especially with rules in place. Blocking a transaction that is not actually fraud damages customer experience and increases friction in the instant payment process. By employing AI models designed to reduce false positives, banks can avoid these negative customer outcomes and continue to grow their business.

Consider this example. A bank implements instant payments, and they soon see an uptick in resulting fraud losses. In response, they implement scalable, real-time AI models to screen and recognize fraudulent patterns of money movement in a matter of milliseconds, blocking them instantly. The models analyze different sets of transaction data points (IP, Country, Bank, Addresses, Currencies, Counterparties, etc.) to recognize and block potential account takeover and impersonation. As a result, the bank reduces fraudulent payments and fraud losses, improves customer experience by blocking fewer legitimate payments, and saves investigator time and resources.

How Hawk AI Technology Prevents Instant Payment Fraud in Real Time

Hawk AI's fraud prevention technology includes pattern and anomaly detection, false positive reduction, and other machine learning models that can prevent fraud in real time and at scale. With its short implementation timelines, proprietary fraud pattern library, and a configurable case management system, banks can use Hawk AI’s technology to prevent instant payment fraud and reduce losses.

To learn more, request a demo or more information. 
 


Share this page