Driving Toward Carbon Neutrality in United States: Do Artificial Intelligence Shocks, Energy Policy Uncertainty, Green Growth, and Regulatory Quality Matter?

The issue of ecological degradation is evolving, presenting new challenges for both human existence and the ecosystem amid rising economic growth. Artificial intelligence (AI), as a transformative technological innovation, brings both opportunities and risks for environmental sustainability. This paper presents empirical evidence regarding the environmental costs and benefits of AI by analyzing its asymmetric impact on carbon emissions in the United States from the first quarter of 1996 to the fourth quarter of 2020. We employ a unique methodology that integrates nonlinear autoregressive distributed lag (NARDL), wavelet time coherence (WTC), and Quantile on Quantile Kernel-Based Regularized Least Squares (QQKRLS) to investigate the influence of AI, energy policy uncertainty (EP), green growth (GG), and regulatory quality (RQ) on achieving carbon neutrality. The research indicates that artificial intelligence (AI) exerts a dual influence on the environment. On one hand, innovations driven by AI enhance energy efficiency and reduce emissions; on the other hand, their high computational requirements and resource consumption contribute to an increase in carbon emissions. Significantly, the adverse effects of AI surpass its positive contributions, leading to a net increase in emissions without effective regulatory oversight. Nevertheless, RQ and GG are crucial in mitigating the negative environmental effects of AI, as regulatory measures can effectively counteract detrimental impacts and enhance positive outcomes. The robustness of these findings is supported by strong correlations identified through WTC and QQKRLS analyses. These results underscore the necessity for proactive regulatory frameworks that aim to optimize the environmental benefits of AI while minimizing its negative externalities, ensuring alignment with the US decarbonization strategy.