Credit Risk Prediction in Unbanked and Microfinance Populations: A PRISMA-Based Systematic Literature Review

This paper presents a Systematic Literature Review (SLR) following PRISMA guidelines, focused on credit risk prediction in unbanked and microfinance populations. The review connects methods from artificial intelligence (AI) and agent-based modeling (ABM) to identify how these approaches have been applied to financial inclusion problems. Using a semi-automated workflow built around a Retrieval-Augmented Generation (RAG) pipeline, 207 studies were screened and evaluated across six inclusion criteria, of which 95 met the established conditions. The results reveal not only clear methodological trends but also recurring weaknesses in model transparency and fairness, highlighting concrete directions for explainable and responsible credit risk modeling.