Is it All Hype? ChatGPT’s Performance and Disruptive Potential in the Accounting and Auditing Industries

Abstract

ChatGPT frequently appears in the media, with many predicting significant disruptions, especially in the fields of accounting and auditing. Yet, prior 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 CPA, CMA, CIA, and EA (enrolled agent) certification exams. We find that the early released ChatGPT 3.5 model is unable to pass any exam (average score across all assessments of 53.1 percent). 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 percent, providing 10-shot training improved scores an additional 6.6 percent, 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 percent. After all these improvements, ChatGPT passed all exams with an average score of 85.1 percent. This high performance indicates that ChatGPT has sufficient capabilities to potentially disrupt the accounting and auditing industries, which we discuss in detail. Lastly, this research provides practical insights for accounting professionals, investors, and stakeholders on how to adapt and mitigate the potential negative consequences of this cutting-edge technology in accounting and auditing firms.

Keywords: Artificial intelligence, chatbots, GPT 3.5, GTP 4.0, Accounting certification, large language model

JEL Classification: M4, M40, M42, M49