Ethical AI Frameworks for Responsible Internal Auditing Practices: A Conceptual and Theoretical Framework

Abstract

The accelerating integration of Artificial Intelligence (AI) into internal auditing is reshaping assurance practices through continuous auditing, advanced analytics, and predictive risk assessment. While AI enhances audit efficiency, coverage, and timeliness, it also introduces significant ethical, accountability, and governance challenges, including algorithmic opacity, data bias, privacy risks, and the potential erosion of professional judgment. Addressing these issues is essential for preserving audit integrity and stakeholder trust in increasingly automated audit environments. This study adopts a conceptual and theory-driven approach, drawing on Stakeholder Theory, Ethical Decision-Making Theory, and Technology Governance Theory. Through a systematic synthesis of academic literature, professional auditing standards, and global ethical AI guidelines, the paper develops the Ethical AI Audit Framework (EAAF). The framework embeds core ethical principles— transparency, fairness, accountability, explainability, privacy, and integrity—across the internal audit lifecycle, from planning to follow-up. The EAAF emphasizes “human-in-command” oversight, highlights governance enablers such as ethical review mechanisms and bias audits, and identifies auditor competence as central to ethical AI outcomes. This study contributes a domain-specific ethical governance model to support responsible, transparent, and trustworthy AI-enabled internal auditing.