Data and code for JCP

Abstract
10 The G7 countries, as global leaders in adopting artificial intelligence (AI), have pledged to enhance ecological quality
11 (EQ) through sustainable AI integration by 2040. This commitment underscores the need to transition from traditional
12 industrial practices to AI-driven solutions that support ecological systems. The purpose of this study is to investigate
13 the asymmetric effects of AI adoption on EQ within the G7 economies over the period 2000m1–2019m12, using an
14 innovative quantile-on-quantile regression (QQR) approach to capture variations in the AI-EQ relationship across
15 different levels of AI adoption. The findings reveal that at the initial stages of AI adoption (low quantiles), the impact
16 on EQ is modestly positive in most G7 countries. This effect increases stronger in the transition phase and becomes
17 significantly beneficial at higher quantiles of AI adoption. Robustness checks using kernel regularized least squares
18 (KRLS), quantile regression (QR), and alternative measures of EQ confirm these results, ensuring the reliability of
19 the conclusions. The study also highlights substantial cross-country differences in the AI-EQ relationship, indicating
20 that tailored policy measures are necessary to maximize the ecological benefits of AI adoption. This research provides
21 insights into how AI can be leveraged for sustainable development in major economies

摘要
10 作为全球人工智能（AI）采纳的领军国家，G7各国承诺通过可持续的AI整合，到2040年提升生态环境质量（EQ）。此承诺凸显了从传统工业实践向支持生态系统的人工智能驱动解决方案转型的必要性。本研究的目的是探讨2000年1月至2019年12月期间，人工智能采纳对G7经济体生态环境质量非对称影响，采用创新的分位数对分位数回归（QQR）方法，捕捉不同人工智能采纳水平下人工智能与生态环境质量关系的变化。研究发现，在人工智能采纳的初始阶段（低分位数），对生态环境质量的影响在大多数G7国家中是适度积极的。这种影响在过渡阶段增强，并在更高的人工智能采纳分位数上变得显著有益。使用核正则化最小二乘法（KRLS）、分位数回归（QR）以及生态环境质量的替代度量方法进行的稳健性检验确认了这些结果，确保了结论的可靠性。研究还突显了各国在人工智能与生态环境质量关系上的显著差异，表明需要制定针对性的政策措施，以最大化人工智能采纳的生态环境效益。本研究为人工智能如何在大国经济中促进可持续发展提供了洞见。