The role of artificial intelligence and green technological innovation in shaping greener and nuclear energy efficiency: Evidence from developed and developing countries using a panel time-varying approach

Artificial intelligence (AI) and green technological innovation have become central to improving the efficiency of renewable and nuclear energy systems, particularly following the rapid acceleration of AI deployment since
2022. This study investigates the dynamic and heterogeneous effects of AI and green technological innovation, proxied by green patents, on the efficiency of renewable and nuclear energy across advanced and developing
countries over the period 2000–2024. Employing a Time-Varying Interactive Fixed Effects (TV-IFE) model alongside Bootstrap Quantile Regression (BSQR), the analysis captures structural changes, cross-country heterogeneity,
and distribution-specific responses that are not observable using conventional static panel approaches. The results reveal a pronounced strengthening of AI's contribution to energy efficiency in the post-2022
period, reflecting the maturation of AI-driven optimization, automation, and intelligent energy management systems. While green technological innovation initially exhibits a weak or negative effect, its impact becomes
significantly positive as AI adoption deepens, indicating a strong complementary relationship between digital intelligence and green innovation. Importantly, the findings uncover substantial heterogeneity across development
levels: advanced economies experience immediate and persistent efficiency gains from AI, whereas developing economies display delayed but accelerating benefits once technological and institutional thresholds
are reached. The findings offer a forward-looking framework for future research, encouraging scholars to adopt time-varying methodologies, post-AI-boom datasets, and development-specific analyses when assessing the role
of emerging digital technologies in energy transitions.