Factors Influencing the Protection of Geographical Indication Agricultural Products: Towards an Artificial Intelligence-driven Governance Framework

Abstract The protection and promotion of geographical indication (GI) agricultural products have attracted significant global attention due to their economic, cultural, and environmental implications. However, the effective governance of GIs faces persistent challenges, including counterfeit products, inconsistent legal frameworks, and limited consumer awareness. This study aims to identify the factors influencing GI agricultural products protection and their interaction with artificial intelligence (AI) governance mechanisms. Using partial least squares structural equation modelling (PLS-SEM), we revealed significant relationships among key constructs, particularly highlighting the crucial role of institutional frameworks and technological infrastructure in shaping effective GI protection systems. Assessment of the measurement model confirmed the reliability and validity of all seven core constructs, with Composite Reliability values ranging from 0.87 to 0.94, exceeding the commonly used threshold of 0.70. Discriminant validity was successfully established through both Fornell-Larcker criterion and heterotrait-monotrait (HTMT) ratios, with all values satisfying the stringent criteria for construct distinction. The model demonstrated strong explanatory power and predictive relevance, providing a robust framework for understanding how various factors interact to shape effective GI protection systems in the context of AI governance. These findings suggested several strategic recommendations for policymakers, industry stakeholders, and academic researchers.