International payment system Mastercard said it has created its own generative artificial intelligence model, Decision Intelligence Pro, which will help banks evaluate transactions in real time and determine whether they are legitimate. The entire process takes place in just 50 milliseconds.
Ajay Bhalla, president of Mastercard’s cybersecurity and intelligence business unit, told CNBC that the new solution is a proprietary recurrent neural network built from the ground up by Mastercard’s cybersecurity and fraud teams.
“We use morphing models that essentially help harness the power of generative AI,” Bhalla said. “It was all created in-house, we have all kinds of data from the ecosystem. Due to the nature of our business, we see all the transaction data that comes to us from the ecosystem.”
Mastercard’s Decision Intelligence Pro solution is based on a patented recurrent neural network, a fundamental component of generative artificial intelligence, developed entirely in-house. This patented AI solution was carefully developed by Mastercard’s cybersecurity and anti-fraud experts, leveraging the transformative power of generative AI through transformative models. Although open source technologies are used from time to time, most of the AI model is homegrown.
Mastercard’s AI model then takes a unique approach, generating paths through its network similar to a thermal radar to produce the final result. A higher score corresponds to the expected behavior of the cardholder, and a lower score indicates deviation from the norm. Notably, this entire process takes just 50 milliseconds, making it extremely effective for real-time fraud detection.
Mastercard says it has invested more than $7 billion in cybersecurity and artificial intelligence technologies over the past five years.
This includes a number of acquisitions, including a deal to buy Swedish cybersecurity firm Baffin Bay Networks in March 2023.