The next generation of AI will transform the way we protect payments

By Rajat Taneja, President of Technology & Paul Fabara, Chief Risk Officer at Visa

October 23, 2023

Today, the societal benefits of artificial intelligence (AI) are often overshadowed by some of the scarier headlines about the downside risks. While we grapple with the emerging future—a future that will at least in part be shaped by this technology—it’s important to maintain a balanced perspective.

In many cases, AI has already been used as a force for good. It has helped radiologists spot cancer better than they can on their own, helped organizations run more efficiently, and can even help predict what someone will type in an email. At Visa, AI helps us detect and prevent fraud, with our AI-driven fraud-fighting solutions preventing an estimated $27 billion in 2022 alone. We’ve been using AI for 30 years as a foundational tool for delivering secure, convenient, and reliable payments. And over the next 30 years, we believe AI has the potential to transform payments in new and exciting ways.

AI as the foundation for fraud detection  

In payments, AI has been used for decades to help detect and prevent fraud. But, as transactions increasingly move online for greater ease and convenience, they become bigger targets for the growing business of fraud. Fraudsters now have easier access than ever to fraud-enabling tools. For instance, in 2022, fraud cost businesses $41 billion globally—a number that is expected to jump 17 percent in 2023.

And fraud itself is changing. Synthetic identities, ransomware-as-a-service and generative AI-powered social engineering have brought entirely new forms of fraud to the forefront. On top of this, consumers increasingly demand a more seamless payments experience. Half of Americans are less likely to buy something if the checkout process takes longer than 30 seconds.This often leaves businesses with a difficult choice: an easy checkout experience, or top-notch security protocols.

As fraud continues to evolve, the tools and resources we need to anticipate and prevent fraud must advance even faster. AI is one of these critical tools. It can power complex, real-time decisions across the entire payment flow‑—from helping financial institutions reduce downtimes, to making instant purchases more secure. Perhaps most importantly, AI can continuously monitor transactions and customer behaviors to identify unusual patterns, connect these signals across transactions and detect suspicious activity—flagging abnormalities in real time, and in some cases, putting a hold on a potentially fraudulent transaction and escalating it for human review quickly and efficiently. For example, Visa already uses Deep Neural Net (DNN) models to run in real time, scoring real time transactions for fraud and analyzing several events daily to catch and prevent cyber threats.

The next wave of tech brings more fraud-fighting potential

Technology has consistently evolved to meet or even advance the changing needs of society. In recent history, it’s helpful to think of this in four waves.

The first wave brought us the personal computer (PC). Then came the internet, which gave the PC even more power and value. Next, mobile devices gave people the power of the internet on the go.

Now, we’re experiencing the beginning of a fourth wave: large language models (LLMs), or generative AI, which could be the most transformative wave of all.

Beyond understanding and generating human-like text and images, perhaps one of the most exciting future applications of generative AI is in fighting fraud—specifically, in the opportunity to analyze and learn from an unprecedented amount of data compared to traditional predictive models.

While bad actors may attempt to use these tools for fraud, the potential applications of this emerging technology nevertheless present an exciting opportunity for financial services companies. The next generation of AI has the potential to take insights from multiple domains and generate new insights, helping to train fraud tools to make more informed and accurate decisions to differentiate fraudulent transactions from legitimate purchases. With the necessary due diligence, governance and development, generative AI is an opportunity to enable more secure payments and thwart fraudsters by staying one step ahead of the technology.

By building AI into business operations, organizations can better fight fraud and thwart cyberattacks, helping to maintain the integrity of the payments ecosystem. Thirty years ago, Visa pioneered the use of AI in payments. Today, we strive to not only track and address emerging fraud trends, but to constantly evaluate our technology and solutions, with the goal of creating a more seamless and secure payments experience for everyone—whether this is at a network level, or how we can help our clients deploy AI to better protect their businesses.

Above all, establishing consumer trust around AI, technology and data is vital for sustainable innovation and growth. We are proud of the three decades that Visa has been committed to being a trusted partner for AI in financial services, and prioritizing best practices around responsible data stewardship.

If the last 30 years of AI in payments is any indication, we have every reason to be optimistic about future fraud-fighting capabilities—and how the next generation of AI will transform the way we protect payments.

Learn more about how Visa is protecting transactions at all points of the payments journey here.

The content is paid for and supplied by advertiser. The Washington Post was not involved in the creation of this content.

ncG1vNJzZmivp6x7uK3SoaCnn6Sku7G70q1lnKedZLCzscCtoK%2Bdl6e8tryOr6CsmV%2BptaZ5zZ6vrWWXmrumvsCtoKimXaSzbq3IZq6ipJxiwbOtzaydqKqdYsGpsYywmLJlp5p6sb7OrZycrF2lrrq5xKerrGc%3D