As payments become more digital and settle faster, fraud is becoming
The payment fintechs join the bank's free artificial intelligence-driven fraud protection service, adding more detailed information to help spot bad transactions without creating extra hassle for consumers.
"If we can get better data, we can deliver better outcomes," said Jon Borman, head of fraud strategy at Capital One. "Adyen and Stripe are [merchant acquirers] and payment processors in the U.S. and globally. They can provide more data for the transactions that occur through their platforms."
Capital One is among the financial institutions that are applying AI to fraud protection.
The fraud threat is growing, necessitating the need for new technology. Eighty percent of companies were targets of payment fraud or attacks in 2023 — a 15% jump from 2022, according to the 2024
"Collaboration has become essential to fight fraud," said Zil Bareisis, a director at Celent. "No single company is large enough to see all the traffic that is relevant to making a precise decision in real time."
"There is still not enough data transferring between issuers and acquirers," said Don Apgar, director of the merchant payments practice at Javelin Strategy & Research. Since merchants have more data about payments than issuers, Capital One is looking for as many data sources as possible, Apgar said.
While AI is designed to quickly analyze data for clues to potential financial crime such as payments fraud, the amount of information that can be gleaned from a payment is still limited, particularly for e-commerce transactions, Borman said. "You get the name of a merchant, a dollar amount and a category, something like sporting goods of women's fashion," Borman said.
Since Stripe and Adyen sell e-commerce and online storefront technology to thousands of merchants, they can get more granular information about payments faster. With this data, Capital One can spot fraud earlier and reduce extra verification steps or false declines.
"There is still significant room to improve payment performance, particularly in combating fraud and reducing false declines," said Trevor Nies, senior vice president and global head of digital for Adyen, adding that the payment company relies on collaboration with card networks and issuers to expand security risk management. Adyen also uses what it calls "smart messaging," which adapts consumer preference and AI to improve approval rates.
There may be even better fraud-fighting tools on the horizon. The growth of generative AI can create original content out of existing data and is a potential tool to fight fraud by producing scenarios based on incomplete data. But that technology is not yet advanced enough to use for real-time, real-world payment fraud detection, according to Borman.
"By partnering and sharing risk signals, each company can strengthen and improve their own models," Bareisis said. "While firms are typically reluctant to share anything that offers them a competitive advantage, reducing fraud across the ecosystem delivers better outcomes for all."
Capital One may also add more users through its planned
Capital One plans to add other merchant acquirers to build a network for Direct Data Share.
The bank has saved about $1 billion by avoiding false declines through its anti-fraud platform since the platform's launch in 2020, Borman said, but he noted that "most of that" has come in the past few months during testing with Stripe and Adyen.
"We know this works, so our goal is to scale this," Borman said.
