THE IMPACT OF ARTIFICIAL INTELLIGENCE ON AUDITING: A STUDY OF MACHINE LEARNING ALGORITHM IN FRAUD DETECTION

ABSTRACT

The major purpose of this study is to ascertain the impact of Artificial Intelligence on auditing: a study of machine learning algorithm in fraud detection. The integration of AI and ML algorithm in auditing firms where the main focus of this study. The population of this study included 150 professionals from 50 auditing firms and financial institutions. Descriptive analysis was used to provide an understanding to how widely AI and ML are being adopted within auditing and Benford’s law was applied to detect anomalies in financial data using AI models trained to analyze digit frequency distributions. The major findings of the study were: (1) The study confirms that AI-powered tools significantly outperform traditional auditing methods in fraud detection accuracy.  (2)  While AI adoption is moderate (mean = 0.93/5), challenges such as low training comprehensiveness (mean = 0.44/5) and organizational barriers in large firms (regression coefficient = -1.24) highlight systemic gaps. (3) The negative impact of investigation costs (coefficient = -0.26) and firm size on AI effectiveness signals the need for adaptive regulations. The recommendations proposed where as follows: (1) Given AI’s superior accuracy in detecting fraud, stakeholders should prioritize refining AI algorithms to handle sector-specific risks, such as invoice fraud in manufacturing or money laundering in banking. (2) The moderate AI adoption and low training comprehensiveness highlight the need for targeted education. (3) Adoption    To mitigate barriers like investigation costs and organizational resistance in large firms , policymakers should introduce tax incentives for SMEs adopting AI tools and streamline compliance requirements for AI-driven audits.