Text Mining of Key Audit Matters–Analysis By Using AI Application

<p>The objectives of this paper are: (1) Text mining of Key Audit Matters (hereafter “KAM”) (2) Analysis of KAM by a logistic regression model Conclusions are: (1) There are significant differences in the words used in KAM reports between Big 4 auditing firms and small and medium accounting firms. (2) In the Toshiba scandal of 2015, its investigation committee pointed out several fraudulent accounting procedures. After the Toshiba scandal, Big 4 auditing firms were sensitive to a keyword “construction” and the client’s judgment. (3) Small and medium accounting firms are very sensitive to accounting fraud and they used the word “fraud” in their KAM reports. (4) The probability of Big 4 auditing firms using “construction”was much higher than small and medium accounting firms. On the contrary, small and medium accounting firms tended to use keyword “fraud” in theirKAMreports. (5)While AI-powered text mining in KAMand financial statement analysis by CPAs offers great potential to improve efficiency and uncover valuable insights, addressing ethical issues is crucial. As AI technology evolves, integrating ethical principles into its application is necessary.</p> <p> </p>