Risiko bias algoritma dalam credit scoring berbasis ai pada bank syariah

Abstract

The use of artificial intelligence (AI) in credit scoring at Islamic banks offers significant efficiency, but on the other hand poses the risk of algorithmic bias that conflicts with the principle of fairness in Islamic finance. This bias generally stems from unrepresentative historical data and opaque algorithm structures, which have the potential to reinforce social injustice and hinder financial inclusion. This study aims to analyze the forms of algorithmic bias in Islamic financing systems, evaluate their impact on the principle of fairness, and offer mitigation strategies based on maqāṣid al-sharī‘ah. The method used is a qualitative literature review with relevant literature search. The analysis results show that algorithmic bias can lead to discrimination against vulnerable groups, reduce public trust, and undermine the values of maqāṣid al-sharī‘ah. Therefore, the proposed mitigation strategies include ethical and inclusive algorithm design, sharia-based algorithmic audits, and the application of explainable AI to ensure system transparency and accountability. The integration of maqāṣid principles into AI governance is key to ensuring that technological innovation remains aligned with the values of justice and inclusivity in Islamic banking. In conclusion, strengthening regulations, ethical governance, and sharia oversight are essential to prevent and minimize the risk of algorithmic bias in AI credit scoring systems at Islamic banks.