How Mastercard Stopped $13B in Fraud With Real-Time AI

Nirav Mehta, CTO of Mastercard Services, explains how Mastercard turned its payments and partner data into a governed enterprise platform for real-time insights, fraud prevention, and new products. He details how AI powers Decision Intelligence Pro, which has prevented about $13 billion in fraud this year, and why CIOs should rethink data architecture, talent, and agent-to-agent commerce.

Key points

• Mastercard Services is a $14 billion business, growing about 20% year over year, and Mehta leads engineering, analytics, and decisioning.
• The data commercialization platform unified previously siloed data with governance, privacy, security, and interoperability across cloud providers.
• A location-data use case cut freshness from 8–10 weeks to hours, and sometimes under 10 minutes, for customer consumption.
• Decision Intelligence Pro uses hundreds of AI models at authorization time and has prevented about $13 billion in fraud this year.
• Mehta says AI is shifting hiring toward systems thinking, critical thinking, and product-design skills, not just coding.
• Agent-to-agent commerce is emerging, but identity, intent verification, security, and standards remain unsettled.