•  
  •  
 

Abstract

Is the Earth saved or sacrificed by artificial intelligence? Our inquiry into whether AI innovation improves environmental sustainability, as assessed by the Load Capacity Factor (LCF), a comprehensive indicator that includes both ecological supply and demand, is centred on this subject. Study uses moderation analysis to investigate synergistic effects using Driscoll-Kraay standard errors to address cross-sectional dependence and heteroskedasticity using panel data from 20 countries between 2017 and 2025. According to our research, LCF is greatly enhanced by AI patent filings, and this benefit is further reinforced by institutional quality and FDI. Crucially, the connection between AI and skill-biased technological transformation increases environmental advantages, indicating that advanced technological capabilities and skilled labour are necessary to realise AI's sustainability potential. However, while renewable energy exhibits conditional significance, urbanisation continues to pose a danger to environmental quality. The dangers of model misspecification in cross-sectionally dependent panels are highlighted by the startlingly conflicting negative AI coefficients produced by traditional FGLS estimation. Our findings refute oversimplified theories on AI's environmental impact by showing that the technology is neither intrinsically green nor detrimental; rather, its effects rely on institutional frameworks and complementary technological abilities. These findings have immediate policy ramifications for accomplishing SDG 7 (cheap, clean energy), highlighting the need to combine AI investments with strong governance and the development of human capital.

First Page

5

Last Page

18

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Download

Share

COinS