The Importance of Auditing Artificial Intelligence: A Governance, Risk, and Assurance Perspective

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Artificial intelligence (AI) has moved from experimental deployment to embedded infrastructure across finance, operations, healthcare, retail, and public administration. As organizations delegate consequential decisions — credit scoring, fraud detection, hiring, medical triage, dynamic pricing, and regulatory reporting — to machine learning models and generative AI systems, the assurance function faces a widening gap between what these systems do and what boards, regulators, and stakeholders can verify. This paper argues that auditing AI is no longer a specialized niche but a core extension of internal audit's mandate to provide independent, objective assurance over governance, risk management, and control. It examines the drivers behind AI audit — financial, regulatory, ethical, and reputational — surveys the emerging standards landscape (IIA's Global Technology Audit Guide series, ISO/IEC 42001, NIST AI RMF, and the EU AI Act), and proposes a practical audit approach spanning governance, data, model, and monitoring layers. The paper concludes that organizations which fail to build AI-specific audit capability expose themselves to undetected model drift, algorithmic bias, regulatory non-compliance, and erosion of stakeholder trust — risks that are financially and reputationally material.