Finastra: Where AI Is Delivering Value in Payments

Artificial intelligence in banking is often discussed in broad, strategic terms. But in payments, the real story is becoming far more practical.

In this conversation, Radha Suvarna, Chief Product Officer for Payments at Finastra, outlines where AI is already delivering measurable value — not as a future concept, but as a set of tangible use cases embedded into everyday operations.

The first area is fraud detection.

While fraud detection has long been a core capability in payments, AI is significantly improving its effectiveness. Advanced models are now better at identifying anomalies in transaction patterns, helping banks detect fraud earlier and reduce losses. Just as importantly, these improvements are enhancing customer experience. Fewer legitimate transactions are incorrectly flagged, reducing unnecessary friction for users.

Closely linked is sanction screening.

This has always been a regulatory requirement, but it has historically created challenges — particularly around false positives. When legitimate payments are flagged incorrectly, customers are forced through time-consuming investigations. AI models are now helping to reduce these false positives by improving accuracy, ensuring that genuine transactions are processed smoothly while maintaining compliance standards.

A third key area is operations, particularly in payment processing.

Even in highly automated environments, a small percentage of transactions still require manual intervention — often referred to as payment repair. These cases typically involve data mismatches or missing information, requiring operations teams to investigate and resolve issues manually.

AI is now streamlining this process.

Using conversational interfaces and historical data analysis, operations teams can identify the root cause of issues more quickly and receive suggested resolutions. What previously took several minutes per transaction can now be reduced significantly, improving efficiency and enabling faster responses to customers.

These use cases are becoming increasingly important as the payments landscape evolves.

The rise of real-time and immediate payments is removing the margin for delay. Transactions must be processed within seconds, leaving no room for manual review cycles. As a result, AI-driven models are becoming essential to maintaining both speed and accuracy.

At the same time, AI is playing a role in system resilience.

By detecting anomalies in system performance before failures occur, AI can help prevent outages and improve overall availability. This proactive approach allows banks to address issues before they impact transactions, supporting the always-on expectations of modern payment systems.

What emerges from these examples is a clear shift.

The value of AI in payments is not defined by whether it fits a specific category, such as generative AI. Instead, it is measured by its ability to deliver practical improvements — reducing costs, improving efficiency, and enhancing customer experience.

In payments, AI is no longer experimental.

It is operational.