In recent years, the economic impact and environmental contribution of Artificial Intelligence (AI) have gradually become a new focus in academia. This study uses a panel data sample of 50 countries to explore the impact of AI on energy transition (ET), aiming to fill an important research gap. The results highlight several critical insights. First, AI has had a significant positive impact on facilitating the ET. This conclusion still holds after a series of robustness tests. Second, AI positively affects ET by promoting renewable energy technology innovation and upgrading the electricity structure, resulting in both technological and structural effects. Third, the impact of AI on ET is non-linear. Threshold effect models show that AI impacts ET differently at various levels of regulation quality (RQ), exhibiting a double threshold effect. AI hinders ET when RQ is lower than the first threshold value. When RQ is in the second range, AI significantly facilitates ET. However, when RQ exceeds the second threshold value, AI hinders ET again. These findings provide insights into the mechanisms of AI's impact on ET and emphasize that an appropriate level of regulation is crucial for AI to facilitate ET. Finally, this study analyzes heterogeneity and offers targeted policy recommendations.

中文翻译：

---

人工智能对能源转型的影响：监管质量的作用是护栏，而不是一堵墙


近年来，人工智能 （AI） 的经济影响和环境贡献逐渐成为学术界的新焦点。本研究使用 50 个国家/地区的面板数据样本来探索人工智能对能源转型 （ET） 的影响，旨在填补重要的研究空白。结果突出了几个关键见解。首先，人工智能对促进 ET 产生了重大的积极影响。经过一系列稳健性测试后，这一结论仍然成立。其次，人工智能通过促进可再生能源技术创新和升级电力结构，对 ET 产生积极影响，产生技术和结构效应。第三，AI 对 ET 的影响是非线性的。阈值效应模型显示，AI 在不同水平的调节质量 （RQ） 对 ET 的影响不同，表现出双阈值效应。当 RQ 低于第一个阈值时，AI 会阻碍 ET。当 RQ 处于第二个范围内时，AI 显着促进了 ET。但是，当 RQ 超过第二个阈值时，AI 会再次阻碍 ET。这些发现提供了对 AI 对 ET 影响机制的见解，并强调适当水平的监管对于 AI 促进 ET 至关重要。最后，本研究分析了异质性并提供了有针对性的政策建议。