Governance Pathways to Sustainable Development: A Global Analysis of Efficiency, Integrity, and Risk Based on Double Machine Learning and Genetic Algorithm Models

ABSTRACT Achieving the Sustainable Development Goals ultimately depends on how effectively states are governed. Using panel data for 41 countries from 2010 to 2023, this study examines how government efficiency, corruption control, and political risk influence sustainable development, and whether governance digitalization moderates these relationships. Three main conclusions were drawn from the use of double machine learning algorithms. First, government efficiency consistently enhances sustainability performance. Second, effective corruption control promotes sustainable development. Third, political risk has the most significant negative impact on sustainable development; compared to developed countries, developing countries are more vulnerable to the shocks of political risk. Furthermore, artificial intelligence technology has a dual moderating effect: it enhances the positive impacts of government efficiency and corruption control while mitigating the adverse effects of political risk. The digital transformation of governance thus amplifies the benefits of efficiency and cushions economies against external risks, underscoring its complementary role in sustainable development. Overall, this study provides global empirical evidence on how artificial intelligence technology is reshaping the institutional foundations of sustainability.