Explainable Artificial Intelligence in Finance: Mapping Business Problems and Technological Solutions

• Systematic review of 35 studies on explainable AI in finance • Common financial XAI use cases: credit risk, market forecasting, fraud detection. • Most frequently used XAI methods in these domains: SHAP and LIME • XAI benefits: increased transparency, stakeholder trust, and regulatory compliance. • Developed taxonomies for business problems, XAI methods, and financial use cases. The rapid adoption of Artificial Intelligence (AI) in the financial sector has transformed key areas such as credit risk management, capital markets forecasting, and fraud detection. Nevertheless, the mounting reliance on complex, opaque AI models, often referred to as black boxes, has raised critical concerns regarding transparency, accountability, fairness, and regulatory compliance. This research addresses these challenges by systematically mapping business problems in finance to technological solutions offered by explainable artificial intelligence (XAI). This study uses the Problem Space Mapping Framework (PSMF) and a systematic literature review to identify and analyze 35 representative financial XAI (FinXAI) use cases. From these use cases, business problems and XAI methods are then extracted and categorized. The results of the study include three comprehensive taxonomies that structure the dimensions of business needs, technological capabilities, and their intersection in practical use cases. The findings indicate that machine-learning-based methods such as SHAP and LIME are predominantly employed in the domains of credit risk management, capital markets forecasting, and fraud detection. These methods enhance process transparency, stakeholder trust, and regulatory compliance. This research offers actionable guidance for decision-makers in the financial industry. It helps them to better understand, select, and implement XAI technologies to address the pressing need for explainability and responsible AI use in high-stakes financial applications. The discussion encompasses limitations and prospective avenues for future research, underscoring the necessity for persistent development and empirical validation of XAI frameworks in the domain of finance.