Can artificial intelligence mitigate environmental inequality? Evidence from leading robotic-driven economies using quantile-based methods

摘要

The rapid integration of AI into global economies raises critical questions about its environmental impact, especially concerning the equitable distribution of carbon emissions. This aspect of climate justice remains largely unexplored. Existing literature extensively examines AI's aggregate environmental impacts; however, it largely overlooks the differential effects across emission quantiles and their implications for environmental inequality. This oversight creates a significant research gap. This study addresses this critical void by investigating the dynamic relationship between AI adoption and carbon emissions inequality (CEI) across ten technologically advanced economies from 2000 to 2023, employing a novel multivariate quantile-on-quantile regression (MQQR) methodology that captures nonlinear dependencies and heterogeneous effects across the entire distribution of environmental inequality. The empirical findings reveal stark heterogeneity across nations and quantile distributions, demonstrating the complex, non-uniform nature of AI's environmental equity implications. European economies exhibit a three-phase trajectory: early AI adoption increases emissions inequality; mid-level adoption stabilizes it; and advanced integration reduces it, particularly in Denmark, France, and Sweden. Germany displays persistent positive effects across all quantiles. Among Asian economies, Japan and Singapore transition from initial increases to reductions in emissions inequality at higher quantiles, while Korea shows consistently positive effects. China presents an oscillating pattern, and the United States exhibits alternating effects across different quantiles. These results underscore the double-edged role of AI in environmental equity, revealing that the relationship between technological advancement and climate justice is neither universal nor linear. The findings provide crucial insights for policymakers, emphasizing the need for quantile-specific, country-tailored AI governance strategies that account for national developmental stages and distributional impacts to ensure AI-driven transformation contributes to environmental justice objectives.