07/07/2025

Data Takes Fraud Spotlight as Banks Shift to Real-Time Operations

PYMNTS

Artificial intelligence has become a buzzword so pervasive in financial services that it may already be at risk of losing its punch.

Behind the hype, the fundamentals of how banks and FinTechs operate are being fundamentally reshaped.

This isn’t a wholesale rug-pull of core infrastructure. Banks and FinTechs have long used AI, even if they don’t always recognize it. Legacy systems built around credit scoring and risk modeling, for example, were early forms of machine learning.

“From my point of view, AI and machine learning have been integrated into these financial institutions for years … the term AI has broadened, though, so there’s more underneath that umbrella now,” Eric Stratman, senior director, analytics and insights, at ValidiFI, told PYMNTS.

What’s changed today isn’t just the tools, but the scale, speed and complexity of the problems they now address, he said.

“In the past, credit risk models were viewed as standalone systems,” Stratman said. “Today, machine learning is being used to identify fraud in real time, validate accounts instantly, and minimize transaction failures, all with little to no human input.”

Still, this new functionality also comes with its own challenges. Chief among them is trust.