Can artificial intelligence reduce energy vulnerability? Evidence from an international perspective

Due to the global energy supply and demand conflicts, climate change, and geopolitical disputes, there is a growing focus on the vulnerability of energy system. Practically, the potential of artificial intelligence (AI) in mitigating energy vulnerability has been noticed. This paper employs panel data spanning 20 years (2000–2019) from 52 countries to examine the relationship and mechanisms between AI and energy vulnerability, and further examines the potential threshold effects and heterogeneity. The findings are fourfold. First, AI emerges as an effective tool in mitigating energy vulnerability. Second, mechanism analyses demonstrate that AI reduces energy vulnerability via improving the government governance effectiveness, promoting green technology innovation and facilitating industrial structure upgrading. Third, threshold effect tests show that when economic growth and financial development reach higher threshold levels, the mitigation effect of AI on energy vulnerability becomes more pronounced. Lastly, heterogeneity results indicate that regions with higher income levels and better digital infrastructure experience a more significant reduction in energy vulnerability due to the full utilization of AI. These findings provide valuable insights for policymakers and energy organizations seeking to reduce energy vulnerability and ensure energy security.