Authors: Krishna Prisad Bajgai, Dr. Bhojraj Ghimire, Niraj Kumar Shah, Netra Prasad Joshi
Abstract: Artificial Intelligence (AI) and Machine Learning (ML) technologies are increasingly applied in financial institutions for credit scoring, fraud detection, algorithmic trading, and risk management. Although these techniques offer high predictive performance, many models operate as complex “black-box” systems whose decision-making processes are difficult to interpret. This lack of transparency creates challenges related to trust, fairness, and regulatory compliance. Explainable Artificial Intelligence (XAI) aims to provide transparency and interpretability to AI-based models by offering explanations for their predictions. This paper explores the role of explainable AI in financial decision systems, focusing on its applications in credit risk assessment, fraud detection, and financial forecasting. The study reviews existing explainability techniques such as SHAP, LIME, and interpretable models, and proposes a conceptual framework for integrating explainable AI into financial decision-making systems. The findings highlight that integrating explainability mechanisms improves trust, transparency, and regulatory compliance while maintaining model performance. The paper concludes with future research directions for developing trustworthy AI-driven financial systems.
DOI: http://doi.org/10.5281/zenodo.20837802
