
AI in Financial Services Podcast How Payments Firms Secure AI While Meeting Regulatory Demands - with Debjit Saha of MoneyGram
12 snips
Jan 12, 2026 Debjit Saha, VP of Engineering & Product for Risk & Compliance at MoneyGram, discusses navigating the complexities of modern payment systems. He shares insights on how fraud sophistication is rising while customers seek seamless experiences. Debjit highlights the importance of data privacy and regulatory compliance, such as GDPR, in shaping AI controls. He also emphasizes the need for explainability in AI, the role of human oversight, and actionable steps leaders can take to enhance credibility in just 90 days.
AI Snips
Chapters
Transcript
Episode notes
Fourfold Pressure On Payments
- Payments teams face simultaneous pressure from regulators, customers demanding low friction, sophisticated fraudsters, and cost constraints.
- These forces require more advanced data enrichment and smarter decisioning rather than only legacy tooling.
Privacy Forces Explainable Models
- Privacy laws like GDPR and RTBF constrain data use and complicate model training across jurisdictions.
- Explainability and auditability become essential because many risk models cannot be black boxes for regulators.
Design Controls Around Human Roles
- Define each control's purpose clearly and map where human judgment must remain in the loop.
- Automate low-value work while keeping humans on high-value decisions and sample quality-control checks.
