Blessing or peril? The impact of Artificial Intelligence on the energy transition—evidence from China

The energy transition is essential for combating climate change and achieving sustainability. Artificial intelligence, as a ‘double-edged sword’ technology, not only provides new momentum for the energy transition but also poses risks of uneven development. This paper investigates the impact of AI on the energy transition using data from 261 cities in China. The findings indicate AI can reduce energy consumption, enhance energy efficiency, and optimize the energy structure, thereby promoting the energy transition across three dimensions: quantity, quality, and structure. The mechanism analysis confirms that AI can indirectly promote the energy transition by improving human-machine alignment and fostering economic agglomeration. Moderation results indicate that the quality of digital infrastructure, technological absorptive capacity, and production factor structure significantly influence the effectiveness of AI. Threshold effects reveal that the impact of AI on the energy transition exhibits an increasing marginal effect under different levels of energy equity. Spatial spillover analysis reveals a significant spatial dependence in the energy transition. However, due to the predominance of the siphoning effect over the trickle-down effect, the impact of AI on neighbouring areas temporarily exhibits a dampening effect. This study enhances the understanding of opportunities presented by AI, providing important references for promoting energy transition and ecological civilization construction.