The Application of Generative AI in Internal Auditing—A case study of DeepSeek
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Under the strategic background of strengthening audit through technology and the construction of audit informatization, this article explores the application of generative artificial intelligence in internal audit. Taking the domestic model DeepSeek as an example, the study systematically analyzes the compatibility of its functional characteristics, such as multi-format document parsing, long text processing, and deep adaptation to Chinese scenarios, with the needs of internal audit work. The article focuses on three stages: pre-audit investigation, audit implementation, and report formation, specifically exploring the application scenarios and empowerment paths of DeepSeek in key areas such as legal system sorting, regulatory information analysis, preliminary evaluation of financial data, intelligent review of textual materials, auxiliary judgment of irregularities, and report generation and proofreading. The research shows that DeepSeek can effectively enhance the breadth, depth, and efficiency of audit work, promoting the transformation of audit mode towards intelligence. However, it is also necessary to pay attention to risks such as data security, model illusion, and over-reliance. Finally, the article proposes suggestions such as building a human-machine collaboration framework and strengthening the cultivation of composite talents, aiming to provide reference for promoting the safe, compliant, and efficient integration of generative artificial intelligence into internal audit practice.
Fields of Study & Subjects
Robotic Process Automation Applications
Artificial Intelligence Applications
Auditing
Earnings Management
Governance
Publication details
- Journal / source
- Academic Journal of Business & Management
- Volume
- 8
- Issue
- 2
- Publication date
- 1 January 2026
- Language
- EN