Artificial Intelligence: A Blessing or a Curse for Climate Action (SDG 13)? The Moderating Roles of Governance Quality and Digital Infrastructure

This study examines the dynamic relationship between the adoption of artificial intelligence (AI) and carbon dioxide (CO2) emissions, focusing on the moderating roles of governance quality (GQI) and digital infrastructure (DII) across 104 countries from 2000 to 2023. Using two-step system GMM and two-stage least squares (2SLS) estimations, the findings reveal that AI, while enhancing innovation and productivity, currently contributes to higher CO2 emissions, particularly in economies with weak governance and underdeveloped digital ecosystems. Strong institutional quality and advanced digital infrastructure significantly mitigate this effect, suggesting that GQI and DII are critical for realizing AI’s potential as a sustainable technology. The results further reveal pronounced heterogeneity across energy-efficient and energy-inefficient countries as well as low-AI and high-AI stages, indicating that the environmental impact of AI is weaker in settings characterized by higher energy efficiency and early-stage AI diffusion, but stronger in energy-inefficient and AI-advanced contexts. These findings underscore the context-dependent nature of AI’s environmental outcomes and highlight the importance of governance-driven digital transformation for achieving sustainable growth.