AI in Banking Industry: Benefit and Issues for Credit Analysis Use Case

Abstract

All banks are increasingly adopting artificial intelligence (AI) to enhance credit analysis. However, deploying AI in credit processes introduces legal and ethical challenges. This study aims to identify and analyze the legal and ethical risks arising from AI-enabled credit analysis in Indonesia and to propose mitigation strategies within an internal governance framework. Through literature review and examination of Indonesian and international regulatory frameworks  the paper discusses four primary risk domains: (1) data privacy and personal data protection when AI consumes extensive customer information; (2) algorithmic bias that may perpetuate or amplify discriminatory lending practices; (3) lack of explainability in complex “black-box” models undermining transparency and regulatory compliance; and (4) cybersecurity vulnerabilities, including adversarial attacks and data poisoning. Findings indicate that robust model risk management, deployment of explainable AI techniques, regular bias testing, stringent privacy impact assessments, and human-in-the-loop review mechanisms are essential to mitigate these risks. The study concludes that by integrating these measures into existing risk and compliance frameworks, Indonesian banks can harness AI’s benefits for credit analysis while maintaining legal compliance and ethical standards.