The Application of Generative Artificial Intelligence for Financial Statement Audit Process of Chiang Mai University Library

Abstract

This research aims to study (1) the approach to apply Generative AI technology, namely Chat GPT and Gemini in the financial statement document auditing process, including the approach to improve the process to be in line with technology, (2) efficiency and effectiveness of technology, and (3) risk management and internal control approaches. This research is qualitative research using quasi-experimental research methods combined with in-depth interviews and observations. The research found that Generative AI can quickly and accurately collect and examine large volumes of financial data, provided the data is prepared in the appropriate format, such as an Excel file using universal date format, it reduces the number of steps in the work process from 6 to 5 steps, reduces the number of steps for data validation from 6 to 3 steps, and reduces the average working time from 19.01 minutes to 10.04–10.45 minutes. Gemini is fast and accurate when data is available, while Chat GPT supports a wide range of data but may be less accurate in some cases. Risks identified include staff skills, acceptance of AI results, data security, and accuracy of results. In conclusion, the use of Generative AI can indeed help improve the efficiency of financial statement auditing, but it requires the preparation of data, personnel, policies, and continuous monitoring and evaluation.