MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE IN CREDIT RISK PREDICTION FOR COMMERCIAL BANKS: A REVIEW

Abstract

Credit risk prediction is a fundamental function of commercial banking, directly influencing financial stability, profitability, and regulatory compliance. With the rapid growth of digital financial data and advances in computational capabilities, machine learning (ML) and artificial intelligence (AI) have emerged as transformative tools in credit risk assessment. This review paper provides a comprehensive synthesis of the existing literature on ML and AI applications in credit risk prediction for commercial banks. It examines the evolution from traditional statistical models to advanced ML techniques, including classical algorithms, ensemble learning methods, and deep learning approaches. The study develops a conceptual framework that integrates data inputs, pre-processing techniques, modelling approaches, decision processes, and governance mechanisms, highlighting the multi-layered nature of AI-driven credit risk systems. A comparative analysis of different models is presented, emphasizing their strengths, limitations, and applicability in various banking contexts. The review also identifies key challenges, including data quality issues, model interpretability, class imbalance, regulatory constraints, and ethical concerns such as bias and fairness. Furthermore, the paper outlines future research directions, including explainable AI, dynamic modelling, alternative data integration, and privacy-preserving techniques. Practical implications for banks and regulators are also discussed, focusing on the need for robust governance and responsible AI adoption. Overall, this review contributes to the literature by providing a structured and critical understanding of how ML and AI are reshaping credit risk prediction in modern commercial banking.