Examining the impact of artificial intelligence on credit risk management: Evidence from Iranian and European banks

Abstract

Credit risk management is one of the most crucial challenges in the banking industry, which can be significantly enhanced through artificial intelligence (AI). This study examines the impact of AI on optimizing credit risk management processes in banks across Iran and Europe. A quantitative-experimental research approach was employed, collecting data on loan applications, credit scores, and customer repayments from 10 major banks in Iran and 10 European banks over the 2016–2024 period. Data Envelopment Analysis (DEA) and artificial neural networks were applied to assess the effectiveness of AI-driven credit scoring models. The findings indicate that banks utilizing AI-based credit assessment models achieve lower default rates and higher prediction accuracy compared to those relying on traditional methods. Moreover, significant differences between Iranian and European banks in AI adoption were observed, highlighting the influence of cultural, economic, and infrastructural factors on technology acceptance. This study provides policy recommendations for banking executives and regulators to enhance credit risk assessment models.