Impact of AI-Based Fraud Prevention on Customer Trust in Digital Payment Systems: An Empirical Investigation of Continuance Intention among Indian Users

Abstract

Abstract The rapid expansion of digital payment ecosystems in India, anchored by the Unified Payments Interface (UPI), has been accompanied by a sharp rise in payment fraud, eroding consumer confidence and threatening long-term adoption. Artificial intelligence (AI)-driven fraud prevention mechanisms—real-time anomaly detection, behavioural biometrics, and adaptive authentication—are increasingly deployed by payment service providers to mitigate this risk. Drawing on the Expectation-Confirmation Model of Information Systems Continuance (Bhattacherjee, 2001) and the trust literature (Mayer, Davis, & Schoorman, 1995), this study empirically examines the influence of perceived AI security features on customers' continuance intention toward digital payment applications. A cross-sectional survey of 350 active digital payment users in India was conducted using a structured, five-point Likert-scale instrument. Pearson correlation analysis revealed a very strong, positive, and statistically significant association between AI security features and continuance intention (r = .900, p < .001, 95% CI [.878, .918]). Simple linear regression confirmed that AI security features significantly predict continuance intention, explaining 76.7% of the variance (R² = .767, F(1, 348) = 1147.00, p < .001; β = .876, t = 33.87, p < .001). The findings substantiate the proposition that visible, intelligent fraud-prevention capabilities function as a salient trust-building cue, thereby reinforcing users' decision to continue using digital payment platforms. The paper discusses theoretical and managerial implications, acknowledges limitations, and outlines directions for future research.