A Comparative Study of Explainable Artificial Intelligence (Xai) Techniques in Financial Auditing Applications
Ganapathy, Venkatasubramanian
Faculty in Auditing Department, Southern India Regional Council of the Institute of Chartered Accountants of India (SIRC of ICAI), Chennai, Tamil Nadu, Bharat
Abstract
The integration of Explainable Artificial Intelligence (XAI) in financial auditing marks a transformative advancement in enhancing transparency, accountability, and trust in automated decision-making processes. This comparative study evaluates various XAI techniques—such as SHAP (SHapley Additive exPlanations), LIME (Local Interpretable Model-agnostic Explanations), decision trees, and counterfactual explanations—within the domain of financial auditing. The findings reveal significant differences in interpretability, accuracy, user comprehension, and auditability across these methods, offering valuable insights for auditors, regulators, and AI developers. The impact of this research is twofold. Firstly, it provides a critical framework for selecting suitable XAI models tailored to specific financial auditing tasks—such as fraud detection, anomaly identification, and risk assessment—thereby improving the reliability of AI-augmented audits. Secondly, the study addresses regulatory and ethical imperatives by demonstrating how transparent AI systems can support compliance with financial standards and accountability norms. Ultimately, this research contributes to the broader adoption of trustworthy AI in finance, promoting more informed decision-making and fostering greater confidence among stakeholders, including auditors, clients, and regulatory bodies. It lays the groundwork for future development of hybrid audit systems that balance AI efficiency with human-centric transparency.
Keywords: Artificial Intelligence, auditability, transparency, XAI techniques
Impact Statement
This comparative study of Explainable AI (XAI) techniques in financial auditing bridges critical gaps between AI transparency and regulatory compliance. By evaluating LIME, SHAP, counterfactuals, rule-based methods, and attention mechanisms across fidelity, interpretability, computational cost, auditor trust, and regulatory alignment, the research provides auditors with actionable guidance for deploying AI responsibly. Key findings reveal:
SHAP excels in regulatory documentation and bias detection but struggles with computational demands. Rule-based systems offer unmatched transparency for policy enforcement but oversimplify complex patterns. Counterfactuals enable actionable remediation insights, while attention mechanisms enhance unstructured data analysis. The framework empowers auditors to select context-optimal XAI methods, strengthening compliance with standards like GDPR, SOX, and Basel III. Hybrid approaches (e.g., SHAP + rule-based) are recommended to balance accuracy and interpretability, fostering stakeholder trust in AI-driven audits. This advances ethical AI adoption in high-stakes financial oversight.
About Author
Mr. Venkatasubramanian Ganapathy, M.Phil., B.Ed., M. Com, D.P.C.S. is serving as a faculty in Auditing Department, Southern India Regional Council of the Institute of Chartered Accountants of India (SIRC of ICAI), Chennai, Tamil Nadu, Bharat. He has over 21+ years’ academic experience and 9 years corporate experience. He has presented and published many research papers in International and National Conferences and journals. His area of interest are Auditing, Finance and Accounting, Taxation, Law, AI, ML, DL, Cloud Computing, IoT, Osmotic Computing, Blockchain Technology, Big Data Analytics, Python, RDBMS, Serverless Computing, Forensic Auditing, Cyber Security, Quantum Computing etc., He has been recognized with many Awards. His focus on implementation of latest technologies in his field.
References
American Accounting Association (AAA). Current issues in auditing. https://publications.aaahq.org/cia/article/18/2/A1/12271/Transparent-AI-in-Auditing-through-Explainable-AI
Research gate. ResearchGate. https://www.researchgate.net/publication/388353445_A_Comprehensive_Comparative_Analysis_of_Explainable_AI_Techniques
Rani, B. T. (2024). Artificial Intelligence tools in Learning English language and Teaching. How can be AI used for Language Learning. Edumania-An International Multidisciplinary Journal, 02(04), 230–234. https://doi.org/10.59231/edumania/9085
Gunjan, G., & Jakhar, M. S. (2024). Studying the computational approaches and algorithms for calculating the generalized commuting probability of finite group. Edumania-An International Multidisciplinary Journal, 02(04), 322–328. https://doi.org/10.59231/edumania/9091
Ganapathy, V. (2024). AI-Based risk assessments in Forensic Auditing: benefits, challenges and future implications. Shodh Sari-An International Multidisciplinary Journal, 03(04), 100–128. https://doi.org/10.59231/sari7750
RESHAPE: Explaining Accounting Anomalies in Financial Statement Audits by enhancing SHapley Additive exPlanations. Cornell University. https://arxiv.org/abs/2209.09157
ScienceDirect (Explainable AI in Auditing. https://www.sciencedirect.com/science/article/abs/pii/S1467089522000240