Comparative analysis of fraud detection using explainable artificial intelligence techniques

Abstract

Artificial intelligence (AI) has revolutionized diverse fields, solving intricate problems. However, as AI advances, its growing opacity raises concerns, potentially eroding user trust and violating transparency laws. The purpose of this work is to thoroughly examine explainable artificial intelligence (XAI) approaches, with a particular emphasis on the difficulties in detecting fraud that arise from employing black-box AI systems. Recognizing the potential consequences of AI’s opacity, the paper emphasizes the risk of user distrust and legal issues arising from non-transparent systems. The narrative gradually shifts to the emergence of XAI as a solution to the black box dilemma. It emphasizes the significance of transparency, accountability and trustworthiness in evolving AI structures. This includes model-specific, post-hoc, and intrinsic interpretability methods. A dedicated literature review is also highlighted, showcasing the practical application of XAI in fraud detection scenarios.