AUDITING THE RELIABILITY OF GENERATIVE AI-GENERATED ACCOUNTING AND FINANCIAL DOCUMENTS: A PROPOSED CONCEPTUAL FRAMEWORK FOR EVALUATING DIGITAL AUDIT EVIDENCE UNDER ISA 500

Wisam Khalaf Nejres

Авторы

Wisam Khalaf Nejres University of Kirkuk, Administration & Economics

Ключевые слова:

Generative AI, audit evidence, document reliability, content provenance, professional skepticism, ISA 500, synthetic documents

Аннотация

This work presents a framework for evaluating the accounting and financial documents generated or significantly altered by generative artificial intelligence (AI) and submitted as audit evidence. 100% human-written – the use of AI itself is not problematic. It is uncertain evidence which creates issues – specifically when it is impossible to determine the origin, input, production history or approval of a document. Through an integrative structured review, normative analysis, and conceptual synthesis, we develop the AI-Generated Document Audit Reliability (AGDAR) framework. The framework monitors the document’s journey through a lifecycle in six linked dimensions (provenance, integrity, traceability, independent corroboration and accountability) consistency. The primary anchors of provenance and independent corroboration; the other dimensions connects knowledge of origin to verification of content and responsibility. AGDAR delineates four levels of AI involvement and correlates these to an incremental audit response. The paper claims that an AI-generated document might not be untrustworthy per se. Strength of the evidence is dependent upon the extent of the production chain that can be verified, and the existence of independent evidence that removes doubt. The framework enables auditors to systematically identify breaks in the chain of evidence, select proportionate procedures and reduce reliance on the appearance of documentary evidence alone.

Библиографические ссылки

H. Li and M. A. Vasarhelyi, “Applying large language models in accounting: A comparative analysis of different methodologies and off-the-shelf examples,” Journal of Emerging Technologies in Accounting, vol. 21, no. 2, pp. 133–152, 2024, doi: 10.2308/JETA-2023-065.

C. Autio, R. Schwartz, J. Dunietz, S. Jain, M. Stanley, E. Tabassi, P. Hall, and K. Roberts, Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile (NIST AI 600-1). Gaithersburg, MD, USA: National Institute of Standards and Technology, 2024, doi: 10.6028/NIST.AI.600-1.

H. Gu, M. Schreyer, K. Moffitt, and M. Vasarhelyi, “Artificial intelligence co-piloted auditing,” International Journal of Accounting Information Systems, vol. 54, Art. no. 100698, 2024, doi: 10.1016/j.accinf.2024.100698.

F. Xiong, Q. Han, and C. Zhang, “Design AI agent for auditing: Applying large language models (LLMs) and retrieval augmented generations (RAG) to audit workflows,” Journal of Emerging Technologies in Accounting, vol. 23, no. 1, pp. 189–198, 2026, doi: 10.2308/JETA-2024-041.

S. Ahmad, A. Harbola, S. Sachdeva, B. K. Yadav, M. K. Sharma, A. Khan, and J. Ahmad, “Generative AI in enterprise IT implications for audit evidence, assurance, and governance,” EDPACS, advance online publication, 2026, doi: 10.1080/07366981.2026.2633846.

A. L. Martinez, “Jurisvision of deepfake financial documents: Semiotic disruption and remediation in post-textual evidence,” International Journal for the Semiotics of Law, advance online publication, 2026, doi: 10.1007/s11196-026-10514-w.

B. Chandra, J. Dunietz, K. Roberts, Y. Lee, P. Fontana, and G. Awad, Reducing Risks Posed by Synthetic Content: An Overview of Technical Approaches to Digital Content Transparency (NIST AI 100-4). Gaithersburg, MD, USA: National Institute of Standards and Technology, 2024, doi: 10.6028/NIST.AI.100-4.

International Auditing and Assurance Standards Board, “Proposed revisions for audit evidence & risk response: ISA 330, ISA 500 & ISA 520,” Exposure Draft, 2026. [Online]. Available: https://www.iaasb.org/publications/proposed-revisions-audit-evidence-risk-response-isa-330-isa-500-isa-520

Coalition for Content Provenance and Authenticity, “C2PA specifications, version 2.4: Content Credentials,” 2026. [Online]. Available: https://spec.c2pa.org/specifications/specifications/2.4/specs/ContentCredentials.html

G. Xu and C. Yang, “It doesn’t make sense: Professional skepticism as sensemaking,” Current Issues in Auditing, vol. 19, no. 1, pp. P10–P15, 2025, doi: 10.2308/CIIA-2024-001.

Financial Reporting Council, “Generative and agentic AI guidance,” 2026. [Online]. Available: https://www.frc.org.uk/library/standards-codes-policy/audit-assurance-and-ethics/guidance/ai-in-audit/

Public Company Accounting Oversight Board, “Staff update on outreach activities related to the integration of generative artificial intelligence in audits and financial reporting,” 2024. [Online]. Available: https://pcaobus.org/news-events/news-releases/news-release-detail/pcaob-staff-shares-observations-from-outreach-on-use-of-generative-artificial-intelligence-in-audits-and-financial-reporting

N. Nandal, “AI-governed audit assurance: A conceptual framework integrating predictive intelligence, governance controls, and continuous assurance,” EDPACS, advance online publication, 2026, doi: 10.1080/07366981.2026.2667950.

S. Ali, K. Mohiuddin, S. Sachdeva, A. Juyal, S. Praveen, and A. Khan, “Adaptive audit frameworks for autonomous and AI-assisted decision ecosystems,” EDPACS, advance online publication, 2026, doi: 10.1080/07366981.2026.2660882.

International Auditing and Assurance Standards Board, “Technology position,” 2024. [Online]. Available: https://www.iaasb.org/publications/technology-position-statement

M. M. Dong, T. C. Stratopoulos, and V. X. Wang, “A scoping review of ChatGPT research in accounting and finance,” International Journal of Accounting Information Systems, vol. 55, Art. no. 100715, 2024, doi: 10.1016/j.accinf.2024.100715.

T. C. Stratopoulos and V. X. Wang, “Artificial intelligence and accounting research: A framework and agenda,” International Journal of Accounting Information Systems, vol. 56, Art. no. 100760, 2025, doi: 10.1016/j.accinf.2025.100760.