AI framework for assessing the impact of geopolitical risk on global tourism demand

Purpose – This study introduces and illustrates an AI-based framework for assessing the impact of geopolitical risk on global tourism demand. It examines how a tourism-specific risk measure can be combined with heterogeneous economic, spatial, and behavioral indicators. Methodology – The study applies a multi-source quantitative approach using tourism, macroeconomic, geopolitical, spatial, and Google Trends data. A composite Geopolitical Tourism Risk Index (GTRI) is constructed to capture destination-specific exposure. XGBoost serves as the AI analytical tool for processing heterogeneous indicators and identifying nonlinear relationships and interactions. Findings – The illustrative application shows that geopolitical instability is associated with differentiated tourism-demand responses across destinations. The GTRI supports the assessment of risk exposure, tourism-flow redistribution, and delayed demand responses. The XGBoost application demonstrates the feasibility of integrating heterogeneous determinants within a unified analytical framework. Originality of the research – The study introduces a tourism-specific Geopolitical Tourism Risk Index (GTRI) integrated into an AI-based tourism demand forecasting framework. The originality lies in combining geopolitical risk, spatial proximity, tourism dependence, behavioral risk perception, and conventional tourism indicators, while XGBoost functions as a flexible tool for their joint analysis.