Credit Risk Prediction Model Using Artificial Intelligence in Digital Financial Systems

Abstract

This study adopts a qualitative approach to examine how AI-based credit risk models are conceptualized and applied in digital finance, with a focus on emerging economies. The novelty of this study lies in: Exploring the human dimensions of AI adoption, including trust and interpretation by credit practitioners; Highlighting ethical and governance issues such as bias, fairness, and data privacy; Identifying hybrid models that combine AI with human oversight in risk decisions; and Offering policy insights for the responsible integration of AI in digital credit systems. This study uses a qualitative exploratory approach to investigate how Artificial Intelligence (AI) is conceptualized, implemented, and perceived in the context of credit risk prediction in digital financial systems. Through a qualitative, multi-case approach involving interviews and document analysis, it finds that AI offers substantial potential to improve credit risk models, particularly through the use of alternative behavioral data. These findings highlight the need for a balanced credit assessment framework where AI supports, but does not completely replace, human expertise.