Trustworthy Artificial Intelligence in Financial Decision-Making: A Systematic Review of Explainability, Fairness, and Accountability

Abstract

The rapid adoption of artificial intelligence (AI) in financial services has introduced critical concerns regarding the transparency, equity, and governance of algorithmic decision-making. This paper presents a systematic review of 43 peer-reviewed studies published between 2018 and 2024, examining three core dimensions of trustworthy AI in finance: explainability, fairness, and accountability. The review synthesizes findings across credit scoring, fraud detection, risk management, and algorithmic trading domains. Results indicate that post-hoc explainability methods such as SHAP and LIME dominate current implementations, while fairness-aware approaches remain underexplored relative to performance optimization. A persistent trade-off between predictive accuracy and fairness is documented across multiple application contexts. This paper contributes a structured analytical framework and identifies gaps that warrant future investigation under evolving regulatory mandates including the EU AI Act.