Visa is introducing
The initiatives are part of Visa's push against
One new service is Visa Deep Authorization, which uses AI to analyze transaction data within Visa's neural network, generating a risk score for each e-commerce and digital transaction, according to James Mirfin, global head of risk and identity solutions.
The tool, available in North America, can help mitigate many types of fraud including a malicious e-commerce scam called triangulation, which can cost an affected issuer more than $100,000 per incident, he said. In this type of fraud, criminals pose as legitimate merchants to intercept money from consumers, while paying the supplier of goods with a stolen credit card.
Visa has also developed a new card fraud protection service for global issuers that extends to real-time and account-to-account payments, including transactions flowing through peer-to-peer digital wallets and central banks' instant-payment systems.
That service generates a risk score for each transaction a bank sends or receives, and the San Francisco-based card network recently completed a proof-of-concept of the service with several U.K. banks, Mirfin said. Banks participating in PayUK, Britain's instant-payments service, will be
"There were hundreds of millions of pounds lost last year in the U.K. to instant-payments fraud, and our test suggests Visa could help block a significant number of payments associated with frauds and also with so-called user-authorized scams," he said. Visa is also testing the service with domestic-payments networks in other countries.
Visa also is expanding its Visa Advanced Authorization and risk management service to cover non-Visa card payments globally so that organizations can provide a single fraud-detection service for all types of payments.
"Issuers are looking for ways to simplify fraud management for transactions they manage across multiple local and international card networks," Mirfin said, noting that Visa has invested $10 billion over the last five years in advanced AI and other analytical systems to battle fraud.
Visa's newest fraud-detection tools, part of its existing Visa Protect service, add a new layer of surveillance by constantly comparing evolving new fraud threats against historic patterns, according to Mirfin.
"We're using AI with a deep learning recurrent neural network model that took several years to develop," he said, adding that digital transactions have opened the door to new types of fraud and scams.
Visa, along with Mastercard, has been steadily widening the services it offers to financial institutions and merchants as open banking and regulatory pressures alter its opportunities, analysts say.
"As new networks like real-time payments and account-to-account gain adoption, card networks need to continue to create propositions that drive value for all market participants—purchasers, issuers and merchants," said Sara Elinson, fintech and payments M&A leader at EY, the global consulting firm.
The challenge card networks face will be tailoring new products to meet unpredictable demand in a fast-changing industry, she said.
"The card networks have an advantage through existing relationships with banks and merchants, plus brand recognition with consumers. On the other hand, new entrants [to the payments industry] have an existing model to study and they are coming armed with new technology to solve for unmet needs," Elinson said.
Visa's largest 265 customers are each using an average of 22 of the card network's 200 value-added services, the firm said.
