AI-Based Fraud Detection in Digital Payment Systems

Abstract

Recent developments in research highlight the importance of electronic payment methods in contemporary financial practice. Nevertheless, fraudulent transactions, phishing, identity theft, takeovers, card not present, and other irregularities undermine the system’s efficiencyNotably, rule-based systems for detecting fraud are inappropriate in the present circumstances, given the complexity of fraudulent activities. As such, the paper proposes an AI-based method to detecting fraud in the electronic payment system. The technique relies on machine learning, mining, and behavioural analytics algorithms to detect and avert fraudulent transactions. The concept described in the paper can examine the transaction’s features, including amount, frequency, locations, devices, merchants, time, and behavioural patterns, to detect fraud. The framework can be implemented to enhance the performance of digital payment fraud detection and response systemsSure enough, in practice, the designed framework would assess the risk score of the transaction and based on the results, approve, notify, request additional information, defer or reject the payment.