Enhanced Fraud Detection Through Customer-Specific Rules in Model Activation

Abstract

This paper proposes the adoption of a customer-focused rule engine for payment disbursement across multiple channels, offering a distinct approach from current practices where banks heavily invest in fraud detection models and broad rule engines. Unlike existing systems, this approach places a unique emphasis on empowering customers to influence payment acceptance rules based on their preferences, such as location and store code. By allowing customers to dynamically modify these rules, encompassing both positive (payment acceptance) and negative (payment declination) scenarios, the system provides a final confirmation of the customer's behavioral pattern. In cases where none of these customer-defined rules are triggered, the system defaults to the bank's standard fraud models and rule engines, presenting an innovative way to enhance customer experience and involve them actively in the bank's overall fraud solution.