Energy poverty is a global challenge that constrains economic development, jeopardizes people's health, and impedes the improvement of people's lives. Artificial intelligence (AI) could be an important tool to reverse this dilemma. We utilize a panel data covering 64 countries during 2000–2019 to examine AI's impact on energy poverty. The findings reveal that the application of AI effectively alleviates energy poverty. After a series of robustness checks, this conclusion remains valid. Moreover, heterogeneity tests show that AI significantly alleviates energy poverty in high-income countries and lower-middle-income countries, but this positive influence is not found in upper-middle-income countries. Mechanism tests indicate that the application of AI can generate economic effects, driving technological progress and enhancing human capital, thereby easing energy poverty. Further discussion reveals AI's impact on energy poverty is long-lasting. In addition, its effects are more prominent in countries that do not participate in cooperative organizations. Our findings offer a fresh perspective and approach to improving the global state of energy poverty. Based on these conclusions, we provide relevant policy implications.

中文翻译：

---

根据全球证据探究人工智能与能源贫困之间的关系


能源贫困是制约经济发展、危害人民健康、阻碍人民生活改善的全球性挑战。人工智能（AI）可能是扭转这一困境的重要工具。我们利用 2000 年至 2019 年期间涵盖 64 个国家的面板数据来研究人工智能对能源贫困的影响。研究结果表明，人工智能的应用有效缓解了能源贫困。经过一系列稳健性检验，这一结论仍然有效。此外，异质性检验表明，人工智能显着缓解了高收入国家和中低收入国家的能源贫困，但在中高收入国家并未发现这种积极影响。机制测试表明，人工智能的应用可以产生经济效应，带动技术进步，增强人力资本，从而缓解能源贫困。进一步的讨论表明人工智能对能源贫困的影响是持久的。此外，其效果在不参加合作组织的国家更为突出。我们的研究结果为改善全球能源贫困状况提供了新的视角和方法。基于这些结论，我们提供了相关的政策启示。