Can Artificial Intelligence Generate Convincing Accounting Research Articles? An Empirical Investigation

This study investigates whether artificial intelligence (AI) can generate credible accounting research articles. If AI is capable of producing high-quality academic work, the authenticity and reliability of scholarly research could be at risk. Design/methodology/approach – Using ChatGPT, a research paper was generated on a meta-analysis examining the link between sustainability reporting and value relevance. After the initial draft was produced, references were manually inserted based on the citations provided by ChatGPT. The paper was then submitted unchanged for peer review. Findings – The AI-generated paper was of reasonably high quality, receiving two major revisions from independent experts in accounting and finance. While concerns remain about the accuracy of references and the validity of results, there is a possibility that reviewers might deem the paper publishable, as they are not obligated to verify every citation or replicate findings if the methodology appears sound. Originality/value – AI’s role in academic writing is still emerging, and its long-term implications for research integrity remain unclear. This issue is particularly pressing given the rapid advancements in AI technology.