AI-Enabled Multi-Criteria Credit Decision Systems for Regulatory Compliance

The employment of Artificial Intelligence (AI) in the financial decision-making process has brought unparalleled effectiveness and accuracy to credit assessment processes.Nevertheless, concerns about transparency, bias and lack of regulatory compliance have been growing around the growing use of opaque machine learning models.The proposed paper outlines a complete system (including its AI-based models) of multi-criteria credit decision aimed at addressing these challenges by balancing the predictive performance and the explainability and compliance requirements.The suggested framework integrates Multi-Criteria Decision Analysis (MCDA) methods, which are Analytic Hierarchy Process (AHP) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), with state-of-the-art AI models (gradient boosting and interpretable neural networks).An automatized regulatory compliance layer makes sure that the policies, such as the General Data Protection Regulation (GDPR), the Fair Credit Reporting Act (FCRA), and Basel III norms adherence, take place.The methodology will adopt fairness-aware algorithms, explainability modules based on SHAP and LIME, and a compliance audit trail that can be used for real-time monitoring.The system was tested with a real-world financial dataset, showing a 17 per cent increase in decision accuracy under regulatory specifications of fairness score, data transparency, and justification of credit decisions.The outcomes indicate that incorporating MCDA methods leads to improved interpretability of credit outcomes without affecting https://iscsitr.com/journals/ISCSITR-IJSRAIML80 performance.The study extends the emerging area of responsible AI in the finance sector by presenting a new architecture of decision-making that meets the requirements of compliance, fairness, and practical feasibility, making it suitable for implementation within the highly regulated domain of financial operations.