Adaptive risk forecasting model for foreign economic activity in the EAEU state regulation system

In the context of the global reorientation of international logistics routes, the reliable functioning of the customs control system becomes critical for ensuring economic security. However, the application of a discrete risk management system (based on local document verification at the time of declaration) leads to a disproportionate increase in the transaction costs of legal participants in foreign economic activity due to the growth of frozen capital costs. The scientific literature insufficiently addresses the institutional transformation of predictive control models within the Eurasian Economic Union (EAEU). The purpose of the study is to develop the theoretical foundations and structural organization of an adaptive risk forecasting model in foreign economic activity based on the concept of distributed control. The study employs methods of institutional and comparative analysis, as well as the principles of economic decomposition of transaction costs. The information base consists of strategic documents of customs authorities and data on advanced foreign experience in the digitalization of customs administration. As a result, a multi-level institutional and economic model of adaptive risk forecasting is proposed. The transition from discrete control to continuous analysis of the digital footprint of transnational supply chains using multidimensional data processing technologies and centralized data lakes is justified. A processing core is formed based on a hybrid intelligence architecture, distributing analytical tasks between machine learning algorithms (for automated release) and human officers (for complex legal assessment). It is proven that the intellectualization of control processes creates a strong stimulating macroeconomic effect by eliminating frozen capital costs, as well as a regulatory and protective effect, ensuring the defense of the national market against price dumping. The research results can be used in developing a unified digital system for controlling the movement of goods within the EAEU. The limitation of the work is its conceptual and theoretical nature. Prospects for future research involve conducting an econometric assessment of the release of business working capital upon the implementation of predictive control algorithms.