Application of artificial intelligence tools in the financial statement audit process

Abstract

Annotation: The article systematizes theoretical knowledge and develops practical algorithms for the application of artificial intelligence tools in the process of auditing financial statements by identifying key advantages, opportunities, as well as risks and limitations associated with the integration of innovative technologies into audit procedures. Using system analysis and synthesis, information on the evolution of auditing from traditional manual procedures to automated systems is summarized. The author has classified various artificial intelligence tools (machine learning, natural language processing, robotic process automation, computer vision) and determined their practical role and possibilities of use in auditing activities. To demonstrate the practical implementation of artificial intelligence in the audit of financial statements, step-by-step algorithms for the application of anomaly detection and natural language processing systems have been developed, covering the stages of preparation, training, model application and documentation of results. The study confirms that modern auditing is undergoing a revolution driven by the digitalization of the economy, which requires a transition from selective inspections to the analysis of all data (continuous analysis) to increase reliability and accuracy. Four main types of artificial intelligence tools are systematized, each of which performs specific functions: machine learning is used to predict indicators and detect anomalies and atypical operations; natural language processing is used to analyze unstructured text documents in order to detect inconsistencies; robotic process automation is designed to perform routine tasks of data collection and reconciliation; computer vision provides automated processing of scanned primary documents and verification of their authenticity. Artificial intelligence is a necessary tool for ensuring high-quality auditing in conditions of increasing data volumes, which ensures a transition to continuous data analysis and the detection of complex, non-obvious to humans, patterns. Keywords: artificial intelligence, audit, financial reporting, machine learning, natural language processing, robotic process automation, digitalization, end-to-end analysis, anomaly detection, audit risks.