Abstract
This paper examines the application of AI-assisted analytical tools in evaluating the effectiveness of government regional development programs in the Republic of Kazakhstan. The study addresses the limitations of conventional evaluation practices that focus primarily on budget execution and formal performance indicators, providing limited insight into socio-economic effects under conditions of pronounced interregional heterogeneity. The objective is to assess the relationship between government program expenditures and gross regional product (GRP) dynamics across regions, as well as to identify differences in program effectiveness associated with regional characteristics. The analytical framework integrates panel econometric methods with AI-assisted decision-support analytics designed to enhance the interpretation and practical relevance of empirical results. The study employs panel regression models with regional and time fixed effects using official regional statistics for 2018–2024. AI-assisted tools are applied at the post-estimation stage to support scenario-based analysis and explore regional differentiation in program outcomes. The results indicate a positive but heterogeneous association between government programs and regional economic growth, highlighting the role of institutional conditions, program design, and regional economic structure beyond funding volumes alone. The paper contributes by proposing an integrated econometric–AI evaluation framework that supports differentiated, evidence-based regional policy design under heterogeneous territorial conditions. Keywords: government programs, regional development, panel data analysis, policy effectiveness, AI-assisted analytics, scenario analysis, regional heterogeneity, Republic of Kazakhstan.