Visa Wants Fraudsters to Pick Another Business

As payments move deeper into the artificial intelligence economy, fraud is evolving alongside it, adopting the same technologies that power faster, more seamless commerce.

James Mirfin, senior vice president and global head of risk and identity solutions at Visa, said the transformation is fundamental.

“Fraud has become a business, an economy,” he told PYMNTS CEO Karen Webster, adding that technology has shifted criminal activity from ad hoc schemes to coordinated, professionalized operations.

Advanced technology has not only empowered defenders. It has given fraudsters new weapons.

Mirfin pointed to the rise of agents, deepfakes and voice cloning as tools that allow scams to scale faster and appear more convincing. Criminals are automating activities that once required human labor, enabling attacks to run continuously and at massive volume.

Equally concerning is the patience of modern fraud operations.

“These guys are patient,” Mirfin said.

Credentials may be tested, shelved and sold months later, after consumers believe the risk has passed. Fraud is no longer about quick wins but sustained exploitation.

An Asymmetric Battle for Banks

The challenge for banks is structural asymmetry. Fraudsters move quickly because they can. Banks move carefully because they must.

“Banks aren’t notorious for being fast,” Mirfin said.

Legacy systems, regulatory obligations and operational complexity limit how quickly institutions can adapt, even as fraud typologies evolve daily. Criminals, by contrast, can experiment, fail and pivot with little consequence.

The problem has grown more complex as payments choice expands. Consumers now move fluidly between cards, faster payments and other rails, often assuming the same protections apply across all of them. That fragmentation makes intent harder to determine and increases the risk of misclassifying legitimate behavior.

 

False Positives, False Declines and Payments Choice

The result is a flood of false positives and false declines that frustrate consumers, cost merchants sales and erode trust, Mirfin said.

Solving the problem requires connecting signals across the entire transaction life cycle, rather than treating authentication, authorization and provisioning as isolated events, he said.

Fraud must be addressed as an economic system, not just a technical flaw. He described waking up every day focused on how fraud evolves because it never stands still.

The strategic objective is to flip the economics.

“It’s making the cost of trying to enact fraud higher than the payback that they get,” Mirfin said.

That means creating friction for criminals while keeping legitimate transactions flowing.

Visa’s role is to give banks tools that allow them to respond faster without rebuilding their infrastructure, he said. That includes products designed to be easy to configure and update as fraud patterns shift, rather than requiring lengthy development cycles.

Just as important, Visa operates intelligence at the network level. By analyzing activity across transactions globally, Visa can detect emerging fraud patterns early and push those insights downstream to issuers that would otherwise see only a narrow slice of activity, he said.

Visa’s approach is to embed intelligence throughout the payments flow, rather than bolt it on at a single point.

At the front end, Visa uses provisioning scores to help issuers better assess risk when credentials are added to wallets or devices. Tokenization and secure protocols reduce exposure before a transaction even occurs.

During transactions, Visa runs network-level AI models across authentication and authorization, allowing signals from one stage to inform decisions at another. The company can pass authentication insights directly into authorization so issuers can make more informed, automated decisions in real time, Mirfin said.

Tools such as Visa Risk Manager allow banks to operationalize that intelligence quickly, adjusting rules and responses without extensive reengineering. The acquisition of Featurespace adds behavioral analytics that focus on understanding what good customers normally do, complementing Visa’s network breadth with deeper context.

The result is a more connected view of customers across payment types, account changes and transaction events, helping banks reduce false declines while staying ahead of emerging fraud patterns.

Consumers also play a role in the defense.

“If something seems too good to be true, it really is often too good to be true,” Mirfin said.

Not all digital payments offer the same protections, particularly outside card networks, and awareness matters as scams grow more convincing through AI-driven impersonation, he said.

The Network Effect

Asked what fraudsters should worry about most, Mirfin pointed to Visa’s scale and collaboration. He said he would like criminals to see Visa as “the biggest threat to their business,” not because Visa acts alone, but because it operates a network effect.

Fraudsters operate ecosystems of their own. Visa’s strategy is to counter them with a larger, coordinated ecosystem of issuers, merchants, acquirers and partners, making fraud increasingly expensive, complex and unattractive.