Visa Ups Its Data Security Game by Spending $9 Billion In Five Years

Visa Inc. has spent more than $9 billion over the past five years to boost cybersecurity and reduce fraud, according to an announcement the network made in a blog post Tuesday.

Of that amount, $500 million has been spent on artificial intelligence and data infrastructure. As a result of its investment in artificial intelligence, Visa says it can power more than 60 different AI capabilities that can automate many of the tasks Visa clients are doing manually today, the card network says.

Thanks to its ongoing investment in data security, Visa says fraud rates on its network are at historic lows, averaging 7 cents per every $100 transacted. 

As part of its data-security efforts, Visa coordinates in-house hacking attacks against its network to test its vulnerability and learn where holes are that can be exploited by hackers. The payoff from vulnerability testing alone included the prevention of about $31 million in fraud on the network in fiscal year 2021, which ended Sept. 30. 

Other efforts to prevent fraud include triangulating fraud attempts, IP addresses, and GPS data to catch criminals. In addition, Visa says it helped to prevent about $2.2 billion in attempted client fraud through its Visa Account Intelligence tool, which uses artificial intelligence and machine learning to identify fraudulent transactions as they occur.

Tokenization is another tool the network says it is putting resources behind. Tokenization is a technique that converts a consumer’s primary account or card number into a numeric code that acts as a substitute real account or card number. Adoption of tokenization by Visa clients is up 60% year-over-year, Visa says. As a result, fraud on the network has declined 28% while approval rates have increased 2.5%. “Any time these numbers move in the opposite direction is a benefit to merchants [and consumers]” says Dustin White, Visa’s chief data risk officer.

As part of its tokenization strategy, Visa introduced its Cloud Token Framework (CTF), which is designed to enhance security and increase approval rates for card-not-present transactions across multiple payment experiences and devices. 

Use of CTF in Europe, for example, has tripled since the start of 2021, and reduced card-not-present fraud by 28%. As CTF is adopted on a broader basis, Visa anticipates a similar reduction globally. “That is as important for business as it is for trust, as 89 percent of cardholders abandon or reduce use of their credential after a fraud event,” the blog post says. 

Another result from Visa’s data-security efforts is that false positives, or legitimate transactions incorrectly flagged as fraudulent, declined by as much as 30%, the network says. False positives are a costly problem for U.S. merchants, costing them an estimated $50 billion in sales in 2021, according to Atlanta-based payments consultancy CMSPI.

“Our goal is to quickly and clearly separate good transactions from bad ones, White says. “We’ve made data security more of the fabric of how we do business.”