Artificial intelligence in accounting: applications, economic impact, and risks

Abstract

The article analyses the practical use of artificial intelligence in the accounting of Russian organisations. Five key areas are examined: recognition and processing of source documents, automatic classification of business transactions, continuous internal control, forecasting of financial indicators and preparation of management reports. It is shown that the economic outcome of implementation is formed by direct labour savings, shorter closing periods and improved quality of accounting data. The Russian software market and the specifics of the regulatory environment are considered separately, including the transition to machine-readable electronic documents and the operation of automated tax-control systems. Key limitations are outlined: dependence of conclusions on input-data quality, the opacity of neural-network models, information-security requirements and the need to preserve the accountant’s professional judgment. The authors conclude that artificial intelligence should be regarded as a means of strengthening the analytical function of accounting rather than a substitute for the specialist’s professional responsibility.