In an exclusive interview with CNBC, Mastercard revealed its innovative use of generative artificial intelligence (AI) to combat fraud within its payment processing network. Termed Decision Intelligence Pro, this AI, developed in-house with minimal third-party components, has proven highly effective in preventing fraudulent activities.

Mastercard, a key player in facilitating electronic transactions worldwide, aims to enhance security and reliability within its payment network alongside other major payment networks like Visa and American Express. Ajay Bhalla, President of Mastercard's Cyber and Intelligence Business Unit, disclosed that the company has constructed a proprietary recurrent neural network, a critical aspect of generative AI, from scratch.

Utilizing data from over 125 billion annual transactions, Mastercard's AI algorithm predicts fraudulent transactions by analyzing relationships between merchants and cardholders, deviating from traditional text-based models. By leveraging a cardholder's merchant visit history, the algorithm swiftly generates scores indicative of typical or anomalous behavior, ensuring fraud detection within milliseconds. Mastercard's technology boasts significant potential, with reported improvements in fraud detection rates of up to 300 percent in certain cases, fueling optimism for enhanced security in financial transactions globally.