Mastercard Creates A Generative Ai Model To Fight Fraud
Mastercard will launch its generative AI fraud detection tool later this year, boosting its real-time fraud detection technology.
Decision Intelligence Pro (DI Pro) uses a recurrent neural network created by Mastercard’s cybersecurity and anti-fraud teams. Mastercard handles billions of transactions annually to train its fraud detection system.
The algorithm learns from cardholders’ purchases to understand merchant-cardholder relationships.
DI Pro analyzes trillions of data points to predict fraud when a cardholder makes a transaction. Based on merchant visits, it predicts cardholder transactions.
DI Pro can determine whether a transaction is legitimate or fraudulent in under 50 milliseconds.
DI Pro is slated to be implemented later this year, but initial modeling suggests it might increase fraud detection rates by 20% to 300%.
“Generative AI is speeding up and improving our anti-fraud solutions, deterring criminals and protecting banks and customers. Supercharging our system will help us predict the next fraud, creating trust in every encounter, said Mastercard VP of Cyber and Intelligence Ajay Bhalla.
In addition to safeguarding cardholders from fraud, Bhalla said DI Pro will “reduce the number of false positives by more than 85%.”
Fraud detection is a major operating cost for banks. A 20% reduction in fraudulent transactions reduces human and financial resources needed to assess fraud.