AI大模型赋能财务舞弊识别的实践探索  

The Practice Exploration of AI Large Model Enabling Financial Fraud Identification

作　　者：叶钦华 黄世忠[2] 

机构地区：[1]厦门天健咨询有限公司 [2]厦门国家会计学院

出　　处：《中国注册会计师》2025年第3期36-40,F0002,5,共7页The Chinese Certified Public Accountant

基　　金：国家自然科学基金项目“基于利益相关者视角的财务舞弊识别及应用”(72172135)的阶段性研究成果。

摘　　要：本文基于天健财判与DeepSeek的融合应用,介绍AI大模型赋能财务舞弊识别的应用场景,分析应用效果。本文研究表明,AI大模型在智能数据采集和智能人机交互的应用方面,显著提升了财务舞弊识别系统的效能,但在反舞弊规则模型优化的应用方面还有待发展。本文的研究启示是,“人机协同”将是AI时代的反舞弊新模式,AI是反舞弊专家的智能助理而非替代者。AI将对注册会计师行业产生深远的影响,会计师事务所及其注册会计师必须因应AI时代的变革,借助AI技术赋能反舞弊,提高审计工作的技术含量。Based on the fusion application of Tianjian Finance Intelligence and DeepSeek,this paper introduces the application scenarios of AI large model to enable financial fraud identification,and analyzes the application effect.This paper shows that the AI large model significantly improves efficiency of the financial fraud identification system in the application of intelligent data acquisition and intelligent manmachine interaction,but the effect of the application of antifraud rule model optimization is not as good as expected.This paper reveals that"man-machine coordination"will be a new anti-fraud model in the era of AI,and AI is an intelligent assistant rather than a substitute for anti-fraud experts.AI will have a profound impact on the accountancy profession,accounting firms and their CPAs must change in response to the AI era with the help of AI technology to improve the antifraud work and technical content of their auditing work.

关 键 词：AI 大模型 财务舞弊识别 人机协同 智能审计 

分 类 号：F23[经济管理—会计学]