Is it all hype? ChatGPT’s performance and disruptive potential in the accounting and auditing industries

IF 4.8 3区管理学 Q1 BUSINESS, FINANCE

Review of Accounting Studies Pub Date : 2024-06-27 DOI:10.1007/s11142-024-09833-9

Marc Eulerich, Aida Sanatizadeh, Hamid Vakilzadeh, David A. Wood

{"title":"Is it all hype? ChatGPT’s performance and disruptive potential in the accounting and auditing industries","authors":"Marc Eulerich, Aida Sanatizadeh, Hamid Vakilzadeh, David A. Wood","doi":"10.1007/s11142-024-09833-9","DOIUrl":null,"url":null,"abstract":"<p>ChatGPT frequently appears in the media, with many predicting significant disruptions, especially in the fields of accounting and auditing. Yet research has demonstrated relatively poor performance of ChatGPT on student assessment questions. We extend this research to examine whether more recent ChatGPT models and capabilities can pass major accounting certification exams including the Certified Public Accountant (CPA), Certified Management Accountant (CMA), Certified Internal Auditor (CIA), and Enrolled Agent (EA) certification exams. We find that the ChatGPT 3.5 model cannot pass any exam (average score across all assessments of 53.1%). However, with additional enhancements, ChatGPT can pass all sections of each tested exam: moving to the ChatGPT 4 model improved scores by an average of 16.5%, providing 10-shot training improved scores an additional 6.6%, and allowing the model to use reasoning and acting (e.g., allow ChatGPT to use a calculator and other resources) improved scores an additional 8.9%. After all these improvements, ChatGPT passed all exams with an average score of 85.1%. This high performance indicates that ChatGPT has sufficient capabilities to disrupt the accounting and auditing industries, which we discuss in detail. This research provides practical insights for accounting professionals, investors, and stakeholders on how to adapt and mitigate the potential harms of this technology in accounting and auditing firms.</p>","PeriodicalId":48120,"journal":{"name":"Review of Accounting Studies","volume":"342 1","pages":""},"PeriodicalIF":4.8000,"publicationDate":"2024-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Review of Accounting Studies","FirstCategoryId":"91","ListUrlMain":"https://doi.org/10.1007/s11142-024-09833-9","RegionNum":3,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"BUSINESS, FINANCE","Score":null,"Total":0}

引用次数: 0

Abstract

ChatGPT frequently appears in the media, with many predicting significant disruptions, especially in the fields of accounting and auditing. Yet research has demonstrated relatively poor performance of ChatGPT on student assessment questions. We extend this research to examine whether more recent ChatGPT models and capabilities can pass major accounting certification exams including the Certified Public Accountant (CPA), Certified Management Accountant (CMA), Certified Internal Auditor (CIA), and Enrolled Agent (EA) certification exams. We find that the ChatGPT 3.5 model cannot pass any exam (average score across all assessments of 53.1%). However, with additional enhancements, ChatGPT can pass all sections of each tested exam: moving to the ChatGPT 4 model improved scores by an average of 16.5%, providing 10-shot training improved scores an additional 6.6%, and allowing the model to use reasoning and acting (e.g., allow ChatGPT to use a calculator and other resources) improved scores an additional 8.9%. After all these improvements, ChatGPT passed all exams with an average score of 85.1%. This high performance indicates that ChatGPT has sufficient capabilities to disrupt the accounting and auditing industries, which we discuss in detail. This research provides practical insights for accounting professionals, investors, and stakeholders on how to adapt and mitigate the potential harms of this technology in accounting and auditing firms.

Abstract Image

查看原文本刊更多论文

都是炒作吗？ChatGPT 在会计和审计行业的表现和颠覆潜力

ChatGPT 频繁出现在媒体上，许多人预测它将带来巨大的破坏，尤其是在会计和审计领域。然而，研究表明 ChatGPT 在学生评估问题上的表现相对较差。我们扩展了这项研究，考察了最新的 ChatGPT 模型和功能能否通过主要的会计认证考试，包括注册会计师（CPA）、注册管理会计师（CMA）、注册内部审计师（CIA）和注册代理人（EA）认证考试。我们发现 ChatGPT 3.5 模型无法通过任何考试（所有评估的平均得分率为 53.1%）。然而，通过额外的改进，ChatGPT 可以通过每个测试考试的所有部分：转移到 ChatGPT 4 模型后，平均得分提高了 16.5%；提供 10 次训练后，得分提高了 6.6%；允许模型使用推理和行动（例如，允许 ChatGPT 使用计算器和其他资源）后，得分提高了 8.9%。经过所有这些改进后，ChatGPT 以 85.1% 的平均分通过了所有考试。这一优异成绩表明，ChatGPT 有足够的能力颠覆会计和审计行业，我们将对此进行详细讨论。这项研究为会计专业人员、投资者和利益相关者提供了实用的见解，帮助他们了解如何适应和减轻这项技术对会计和审计公司的潜在危害。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Review of Accounting Studies BUSINESS, FINANCE-

CiteScore

7.90

自引率

7.10%

发文量

期刊介绍： Review of Accounting Studies provides an outlet for significant academic research in accounting including theoretical, empirical, and experimental work. The journal is committed to the principle that distinctive scholarship is rigorous. While the editors encourage all forms of research, it must contribute to the discipline of accounting. The Review of Accounting Studies is committed to prompt turnaround on the manuscripts it receives. For the majority of manuscripts the journal will make an accept-reject decision on the first round. Authors will be provided the opportunity to revise accepted manuscripts in response to reviewer and editor comments; however, discretion over such manuscripts resides principally with the authors. An editorial revise and resubmit decision is reserved for new submissions which are not acceptable in their current version, but for which the editor sees a clear path of changes which would make the manuscript publishable. Officially cited as: Rev Account Stud