As fraud becomes more sophisticated, financial institutions cannot rely on better AI models alone. Rohit Chauhan, former Executive Vice President of Artificial Intelligence & Fraud Solutions at Mastercard, opens a three-part CDO Magazine interview series by explaining why data architecture is emerging as the real differentiator in fraud detection. Robert Lutton, Vice President of Sandhill Consultants, leads the discussion.

Chauhan argues that models are only as effective as the data and context they receive. While individual transactions may appear legitimate, connected patterns across accounts, identities, devices, locations, and behaviors can reveal a different picture.

For banks, this means moving beyond transaction-centric systems toward architectures that allow AI to understand fraud as a network problem.

Key discussion points include:

• Architecture over model sophistication: Chauhan explains why applying more advanced AI to the same incomplete data produces only marginal improvements.
• Data as the defensive advantage: With AI now widely accessible to fraudsters, Chauhan says financial institutions must capitalize on the richer proprietary data already within their organizations.
• Fraud as an asymmetric contest: Fraudsters need only one attack to succeed, while institutions are expected to stop every attempt, making speed and context essential.
• From isolated events to connected patterns: Chauhan illustrates how two seemingly legitimate transactions can signal fraud when their timing, location, and other relationships are evaluated together.
• A foundational priority for CDOs: Chauhan urges data and AI leaders to invest early in connected data infrastructure because the strength of enterprise AI will depend on the architecture beneath it.

The conversation continues in the next installment.