Artificial Intelligence, Institutional Quality, and Electricity Network Efficiency: Evidence from the Global Energy Transition

Abstract

Despite growing recognition of artificial intelligence (AI) as a general-purpose technology, limited attention has been paid to its role in improving electricity network efficiency and the institutional conditions under which these effects emerge. This study examines how AI development influences electricity network efficiency, proxied inversely by electricity transmission and distribution losses, and whether institutional quality strengthens this relationship within the global energy transition. Using panel data for 59 countries over 2010-2024, the analysis employs two-way fixed effects, instrumental variable generalized method of moments (IV-GMM), and method-of-moments quantile regression (MM-QR) to address endogeneity and capture distributional heterogeneity. The results show that AI development significantly improves electricity network efficiency by reducing transmission and distribution losses, with stronger effects observed in high-loss environments. Institutional quality further amplifies these efficiency gains, indicating that governance capacity and digital capability operate as complementary inputs. Heterogeneity analysis shows that the efficiency-enhancing effect of AI is more pronounced in fossil-fuel-dominant systems and technologically advanced economies. The findings highlight that technological adoption alone is insufficient; governance alignment is essential for translating digital capability into operational performance gains. This study contributes to the literature by providing cross-country evidence on the joint role of artificial intelligence and institutions in shaping electricity network efficiency and advancing the global energy transition.