ASSESSMENT OF THE EFFECTIVENESS OF AI METHODS IMPLEMENTATION IN THE MANAGEMENT ACCOUNTING SYSTEM

Abstract

The article proposes methodological approaches to assessing the effectiveness of implementing artificial intelligence models in the management accounting system of retail enterprises. The study focuses on the cost component of AI implementation and its accounting support, since the introduction of analytical technologies requires not only technical integration but also a clear mechanism for identifying, grouping and monitoring related costs. It is substantiated that the use of AI models for sales forecasting contributes to improving the quality of managerial decisions, optimizing inventories, reducing losses, preventing stockouts and shifting from reactive to proactive management of business processes. The main groups of costs associated with the implementation of AI models are identified, including costs of analytical personnel, cloud infrastructure, data processing and storage, model development, integration with ERP and BI systems, technical support and model maintenance. A methodological approach to grouping these costs into current and investment components is proposed, with their subsequent reflection in the system  of analytical accounts of management accounting. This approach makes it possible to distinguish the costs of AI implementation from general IT and administrative costs and to create an information basis for further evaluation of economic effects. The article also considers the specific features of assessing the economic effect of AI models, which is often indirect and becomes visible through improved forecasting accuracy, reduced write-offs, lower product shortages, better inventory turnover and faster preparation of management reports. Particular attention is paid to the scaling effect, which arises when AI models are expanded to other product categories, stores or management processes. It is emphasized that the economic benefits of AI implementation may not appear immediately, since they depend on data quality, model adaptation, user involvement and integration of forecasting results into regular management procedures. The practical value of the study lies in developing an approach to more transparent management of costs related to analytical technologies and in supporting managerial decisions on further implementation and scaling of artificial intelligence in the management accounting system of retail enterprises. The proposed approach can be used as a basis for improving internal accounting policies, management reporting and evaluation of the effectiveness of AI-based analytical solutions.