Enhancing Trust in AI: A Comparative Study of Explainable Machine Learning Models for Critical Decision-Making

The integration of explainable artificial intelligence (XAI) in high-stakes decision making systems has become critical for fostering trust and regulatory compliance across industries. This study evaluates explainable machine learning models in f inancial applications, focusing on their ability to balance predictive accuracy with interpretability while addressing emerging challenges in fraud detection and digital transactions [1, 2]. Through comparative analysis of techniques like LIME and SHAP, we demonstrate how model-agnostic explanation frameworks enhance transparency in black-box systems without compromising performance. Our findings reveal that strategic implementation of XAI principles improves human-AI collaboration in sensitive domains, particularly when combined with real-time monitoring approaches [3]. The research contributes actionable insights for developing auditable AI systems that meet evolving regulatory requirements and organizational risk management frameworks.