Artificial Intelligence and Environmental Sustainability: Investigating the AI‐EKC Nexus for SDG 7 and SDG 13

ABSTRACT This study investigates the nonlinear impacts of artificial intelligence (AI) on environmental sustainability across 81 countries from 2000 to 2020. It constructs a composite AI index using an entropy‐based TOPSIS approach and evaluates long‐run associations with panel techniques that accommodate nonstationarity and cross‐sectional dependence. The evidence points to an Environmental Kuznets Curve (EKC) pattern linked to AI. Broader AI use initially raises energy demand and resource consumption, intensifying environmental pressures, but as adoption deepens, it is associated with sizable gains in energy efficiency and stronger integration of renewable energy. The magnitude and timing of these effects vary with income and resource dependence. High‐income economies experience later but larger improvements, while resource‐intensive economies face stronger near‐term pressures. Further analysis shows that countries with higher initial emission levels benefit more rapidly from AI‐enabled environmental improvements. By combining a comparable AI measure with a unified cross‐country and multi‐outcome perspective over a long horizon, this study offers an integrated view of how AI reshapes energy use and environmental pressures. These results highlight the need for differentiated AI governance strategies that expand clean power and grid capacity, strengthen energy management and transparency for compute‐intensive uses, and advance international cooperation to align diffusion with the aims of SDGs 7 and 13.