| Abstract [eng] |
The increased adoption of artificial intelligence in organisational processes has significantly changed the circumstances under which financial reporting is prepared, audited and assured.The high-risk categories of AI systems that raise complex and layered challenges to auditors and are not fully addressed in existing professional standards are known as high-risk.Meanwhile regulatory regimes controlling AI, most notably the European Union AI Act, have started to introduce structured accountability requirements on organisations implementing such systems.The combination of these trends makes it timely, as well as necessary, to consider how the audit profession is positioning itself in regards to high-risk AI in the context of financial reporting.This is the main topicality of the current research. The body of public disclosures produced by the four largest global professional services and audit firms Deloitte, Ernst & Young, KPMG, and PricewaterhouseCoopers, referred to as the Big Four published between 2023 and 2025 and discussing the topic of artificial intelligence in the context of audit practice, financial reporting, and other related governance and assurance issues. The aim of the study is to examine how the Big Four audit firms publicly disclose AI-related risks, governance, and assurance-related issues in financial reporting audits involving high-risk AI systems. To achieve the aim of this project, the following objectives were defined: 1. To map the characteristics of high-risk AI systems to the audit risk assessment. 2. To develop an integrated theoritical framework that combines audit risk model, professional skepticism, and AI governance frameworks. 3. To design and conduct a directed content analysis of Big Four documents published between 2023 and 2025 by developing a corpus table and a codebook. 4. To compare cross-firm communication themes and identify their implications for audit planning and professional guidance in relation to high-risk AI in financial reporting. On this theoretical basis, a directed content analysis is planned and executed with the help of a purpose-designed corpus of Big Four public reports and a systematic codebook comprising five thematic groups: AI Governance and Assurance, AI and Technology Usage in Auditing, Audit Benefits, Audit Risk, and Professional Skepticism. This analysis has shown that AI Governance and Assurance is the most dominant thematic category in all four firms meaning that AI Governance and Assurance is what the Big Four public discourse frames the issue of high-risk AI.AI and Technology Use in Auditing and Audit Benefits are secondarily but substantively significant themes, and Audit Risk is comparatively little explicitly addressed.Professional Skepticism comes out as the least expressedly stated theme throughout the whole corpus.The patterns of cross-firm communication are widely similar, implying a common discursive orientation towards government as opposed to an explicit sceptical challenge. The research paper is a contribution to the growing body of literature on AI auditing by offering an evidence-based systematic analysis of how the Big Four publicly position themselves with regards to high-risk AI in financial reporting.The implications of the findings on standard-setters, regulators and practitioners include developing more robust professional guidance on audit risk assessment and sceptical judgement in AI-intensive environments. |
