Audit Quality in the Age of AI: How Machine Learning Transforms Fraud Detection
Abstract
This study investigates how the adoption of machine learning tools by audit firms affects audit quality and fraud detection rates. Using proprietary data from Big Four audit firms, we document that ML-assisted audits identify 34% more material misstatements compared to traditional audit procedures, with the improvement concentrated in complex transactions and revenue recognition areas.
机器学习辅助审计能够比传统审计程序多识别34%的重大错报，改进主要集中在复杂交易和收入确认领域。AI工具在模式识别和异常检测方面表现尤为突出。
使用四大会计师事务所的专有审计数据，比较ML辅助审计与传统审计在错报检测率、审计效率和审计费用方面的差异，采用倾向得分匹配控制选择偏差。
AI技术正在深刻变革审计行业。ML工具应作为审计师专业判断的补充而非替代，未来审计准则需要适应技术变革。
论文信息
METADATA引用格式
CITATIONBrown, A., Davis, K., Miller, J. (2025). Audit Quality in the Age of AI: How Machine Learning Transforms Fraud Detection. *The Accounting Review*.
相关研究推荐AI-POWERED
基于学科、关键词、方法论、引用关系的多维度智能推荐
Xu Y., Li H., Wang Z. · JF · 2025-07
Chen L., Pelger M., Zhu J. · JF · 2025-12
Kim S., Muhn M., Nikolaev V. · JAR · 2025-12
Liu X., Chen W., Zhao R. · MS · 2025-08
Smith D., Garcia R., Lee C. · JAR · 2025-09
Kraft A., Schwartz R., Weiss I. · JAE · 2025-06