Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{205371,
        author = {Anish Arvind Karne and Shubham Kailas Badhe and Sriniwas Narayanan Vengarai},
        title = {An Explainable AI (XAI) Framework for Human-in-the-Loop Insurance Fraud Adjudication: Bridging Algorithmic Accuracy with Operational Trust},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {6210-6219},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205371},
        abstract = {Insurance fraud constitutes a significant financial burden on the healthcare sector, yet traditional heuristic-based audit systems are increasingly incapable of detecting the non-linear patterns of modern fraudulent claims. While ensemble machine learning methods offer enhanced predictive capabilities, their "black-box" nature and susceptibility to severe class imbalance issues present significant barriers to enterprise adoption in heavily regulated environments. This paper proposes a novel, hybrid Explainable AI (XAI) framework that integrates SMOTE-balanced ensemble classifiers with a deterministic Business Rules Engine to ensure both predictive accuracy and absolute policy compliance. We demonstrate that by embedding real-time SHAP (Shapley Additive explanations) visualizers within an interactive Streamlit-based dashboard, the system effectively bridges the gap between algorithmic probability and human operational trust. Experimental results indicate that this approach achieves a stabilized accuracy of 86.82% while maintaining high Recall (83.20%)—a critical metric for fraud detection. By providing claims adjusters with instantaneous, plain-English justifications for risk alerts, this research provides a scalable, legally defensible, and high-speed triage system that shifts the paradigm from opaque automation to transparent, "Human-in-the-Loop" adjudication.},
        keywords = {Insurance Fraud Detection, Machine Learning, Fraudulent Claims, Supervised Learning, Classification Algorithms, Data Pre-processing, Feature Selection, Imbalanced Data, Predictive Analytics, Risk Management Insurance, Fraud Detection; Explainable AI (XAI); SHAP; Class Imbalance; SMOTE; Ensemble Learning; Hybrid Adjudication; Streamlit; Human-in-the-Loop.},
        month = {June},
        }