By Ray Birch
PURCHASE, N.Y.— Artificial intelligence is reaching into all corners of financial services, and that obviously includes payments. Mastercard is reporting it has introduced a new solution that uses AI to generate a “risk score” for transactions to evaluate the legitimacy of the person on the receiving end of a digital payment.
The move comes in the wake of real-time payments and the official introduction by the Federal Reserve of its FedNow service. While it’s good news for those making or expecting payments, it has also opened the door to fraudsters, making it easier for crooks who impersonate an individual or company to trick a person into sending money and then to escape with the funds.
Known as authorized push payment (APP) fraud, it has been growing markedly in the United Kingdom and it’s expected to become a large issue in the U.S. as real-time payments take greater hold.
The analysis of the new offering from Mastercard is appearing in CUToday.info as part of an ongoing series around artificial intelligence and credit unions.
Building on Insights
In the U.K., Mastercard said its launch of the Consumer Fraud Risk Score involves the use of large-scale payments data to help identify scams before funds leave a victim’s account.
“The tool builds on insights from Mastercard's work with U.K. banks to follow the flow of money mule accounts over the last few years,” explained Johan Gerber, EVP cyber and security products at Mastercard.
According to Gerber, overlaying money trail information with specific analysis factors—such as account names, payment values, payer and payee history, and the payee’s links to accounts associated with scams—gives FIs greater intelligence to intervene in real time and stop a payment before funds are lost.
‘Known Money Mule Patterns’
“Mastercard is providing an instant risk score—within milliseconds—comparing both payer and payee’s active vs. historic vs. known money mule patterns,” Gerber told CUToday.info. “An important component is how our solution is assessing the relationship and ‘neighborhood’ of the recipient account.
“For example, maybe the recipient account has had fraudulent flows into it, or its transferred funds into accounts that have been part of a network of mule accounts which disseminate fraudulent funds,” he continued. “Organized criminals will pass funds through up to 30 accounts within their network to disguise the initial fraud. AI helps here by processing Mastercard’s insights and giving the predictive score to financial institutions.”
Fraudsters Shift Their Focus
Johan Gerber
Gerber reminded that as banking and payments security becomes increasingly advanced, fraudsters have shifted their focus to impersonation tactics.
“Their goal is to convince people and businesses to send them money, thinking the transfer is to a legitimate person or entity, perhaps a friend, a known supplier or to buy what looks like genuine goods online,” said Gerber.
APP fraud now accounts for 40% of U.K. bank fraud losses and estimates project it could cost $4.6 billion in the U.S. and U.K. alone by 2026, Gerber said.
Mastercard is testing the new technology with U.K. financial institutions first, explaining the banking system in the country makes the score easier to deliver. But the company said it does expect the solution will eventually make its way to the U.S.
‘Not a Card Payment Solution’
“This is not a card payment solution. This is trying to help us solve what is a real growing problem around the world as more countries move to a real-time payment networks from accounts to accounts,” said James Thorpe, SVP C&I communications with Mastercard. “We are doing this in the U.K. first. because the U.K. was one of the first markets to move to real-time payments as the standard—15 years ago. The U.K. has a significant APP fraud problem and it's really tough for the banks.”
During 2022, within the U.K. there were 207,372 incidents of APP scams reported, with gross losses of £485 million, reports show.
“The banks came to us probably about five years ago, challenging us to root out these crimes,” said Thorpe. “With these crimes the money travels through anywhere between five and 30 accounts very quickly, through the banking system, to disguise the funds and then withdraw them from the banking system pretty fast, normally either into cryptocurrency or into an overseas account.”
What Has Happened at One Bank
TSB Bank is one of the first U.K. banks to adopt Mastercard’s new tool.
“In just four months, the bank says it has dramatically increased its fraud detection,” said Gerber. “Based on TSB’s results, the amount of scam payments prevented over a year would equate to almost £100 million saved across the U.K., should their performance be mirrored by all banks. Other banks adopting Consumer Fraud Risk are doing so over the course of 2023 and Mastercard is assessing further international markets to scale the solution.
“Ultimately, AI comes into effect when assessing the broader complex relationships between transacting accounts, the accounts they individually associate with as well as the various transaction patterns,” Gerber continued. “The power of AI is used in processing this level of complexity, across the entire U.K. payments system, and delivering real-time decisions with great accuracy to minimize the impact on legitimate transactions.”