Most people assume Mastercard is a card provider. It’s not.
It’s a global payments company that provides the infrastructure and rules for card payments. The technology moves money between consumers, merchants, banks, and governments in over 200 countries and territories.
The company processes billions of transactions each year (159 billion in 2024) and connects 25,000 financial institutions.
At this scale, it means you don’t have tolerance for poor execution because you can’t afford to break the trust built with businesses over the decades.
In today’s AI at the Top, we will learn:
Mastercard’s AI approach
How the company uses AI for fraud detection and security
Mergers and Acquisitions to scale AI implementation
Mastercard’s AI approach
When we study Mastercard’s AI strategy, three things stand out.
1/ Four Pillar AI Framework
All the initiatives at Mastercard solve for at least one of the four pillars:
Safer: The foundation. Mastercard has spent $11 billion since 2018 to improve security and risk management with AI across 150+ billion transactions annually.
Smarter: Analyses transaction data to help banks, merchants, and governments optimise operations. This is where Mastercard shifts from a transaction processor to an intelligence partner.
More Personal: AI tools for partners to deliver personalised experiences for their direct customers.
Stronger: Uses AI to scale ops, cut costs, and upskill 35,000+ employees.
2/ Hub and Spoke Model
It is difficult to oversee every single AI application. It’s also chaotic to leave all applications to their respective departments.
In our previous edition, we learned how J&J began experimenting with a centralised approach across 900+ use cases, then narrowed them to 10-15% by delegating administration to the respective domains.
Mastercard follows a similar approach.
Hub = Centralised AI leadership team led by the Chief AI and Data Officer, Greg Ulrich. They set up strategies and foundations for the company.
Spoke = Business units like personalisation groups, fraud detection teams, etc. They innovate and deploy AI for their specific domains.
This way, the Hub and Spoke Model allows a balance between innovation and speed while meeting centralised standards.
3/ The Data Moat
In every edition, we learn how enterprises have a data advantage, with decades of data on their business and customers. This paves the way to accurate, actionable insights that any new player would struggle to acquire.
Mastercard is no different.
The total value of transactions processed through Mastercard in 2024 was $9.757 trillion. In 2023, it was $9.027 trillion. From each of these transactions, the company captures merchant details, location, amount, device information, and purchase specifics.
This granular data trains models to build strong baselines, generating multiple use cases.
Enjoying this read? Share the smartest part with your followers. Click to Share.
With decades of holistic, large datasets and solid strategy, here are some AI tools and use cases inside Mastercard:
Decision Intelligence (DI): Real-time fraud detection system by Mastercard’s proprietary gen AI models. It scans over a trillion data points and performs risk assessments in milliseconds, boosting fraud detection rates by an average of 20%.
DI Pro: False positives are a major challenge in fraud detection. Merchants lose revenue, and customers get frustrated. DI Pro analyses transactional history in real time and learns with each transaction. This reduced the false positives by more than 85%.
Vocalink Analytics: Delivers a network-level view of money flows across the entire payments ecosystem. Banks don’t have access to global money flows, but Mastercard does. The company helps banks learn from its network’s insights.
Global Treasury Intelligence: Analyses complex payment data and delivers actionable insights to finance departments. Also, helps large businesses optimise procurement.
AI and Advanced analytics consulting: Mastercard’s AI experts help financial institutions and merchants develop their own AI strategies, build predictive models, prototype new solutions, and optimise costs.
Small business AI coach: Gen AI tool created for entrepreneurs. Provides guidance on business planning, digitalisation, and financial literacy.
SessionM: Focuses on loyalty and customer engagement using AI-driven analytics. Helps brands build unified, 360-degree customer profiles
Agent Pay and Agentic Commerce: Ensures AI agents from OpenAI, Google, and Microsoft shop on behalf of consumers by providing essential payment rails.
YOU DECIDE
🌟 Want to be featured in the next issue? Reach out with your best AI use case and we’ll spotlight it.
Mastercard’s Mergers and Acquisitions.
The enterprise is strategic and has trained its proprietary models. But the company has also invested in acquisitions and partnerships to strengthen its AI game, fast.
Let’s look at some of them.
Acquisitions:
Partnerships:
What can we learn from Mastercard’s AI strategy?
We’ll end this edition with one of the quotes from Mastercard’s President and CTO, Ed McLaughlin.
AI implementation is “not if you could do this (task) with AI, but what’s the problem you could never solve that AI will not allow you to do?”
Sources:
Interviews of Chief AI and Data Officer and CTO
Case studies by Klover, Digital Defynd
Multiple sources on Mastercard’s website
YOUR TURN