Explainable Artificial Intelligence (XAI) in Auditing: A Conceptual Framework for Enhancing Auditor Trust and Judgment Quality

Mohammad Jaefari¹

1) Ph.D candidate, Department of Financial Engineering, Mar. C., Islamic Azad University, Maragheh, Iran

IOI Code : AAAG-AAHHBG
Publication : National Conference on intelligent accounting: data analysis in financial processes(bonabaccuconf.ir)
Abstract :
    This paper explores how Explainable Artificial Intelligence (XAI) can be effectively integrated into auditing to improve auditor trust and professional judgment quality. The central research question is: How can explainable AI be aligned with auditing processes to enhance transparency, trust, and compliance with regulatory standards such as ISA 230 and ISA 315? The study aims to develop a conceptual framework, the XAI-Audit Alignment Model (XAAM), that provides auditors with interpretable, documentable, and regulation-compliant AI outputs. Methodologically, this research employs a structured literature review, conceptual synthesis, and gap analysis based on current auditing and AI practices. Findings suggest that while AI tools have significantly improved anomaly detection and transaction analysis, their lack of transparency poses major risks to audit quality and compliance. The XAAM framework addresses these issues by combining transparency tools (e.g., SHAP, LIME) with auditor-focused interfaces and documentation protocols. We conclude that XAI adoption must be intentional, regulation-aware, and auditor-centric. This paper contributes to the intersection of AI and auditing by offering a practical, regulation-aligned roadmap for implementing explainable AI.

Keywords : explainable ai auditing transparency auditor trust anomaly detection judgment quality xai framework