The effect of artificial intelligence on carbon intensity: Evidence from China's industrial sector

Jun Liu,Liang Liu,Yu Qian,Shunfeng Song
DOI: https://doi.org/10.1016/j.seps.2020.101002
2021-01-01
Abstract:<p>Artificial Intelligence (AI) is becoming the engine of a new round of technological revolution and industrial transformation; as such, it has attracted much attention of scholars in recent years. Surprisingly, scarce studies have shed lights on the effects of AI on the environment, especially with respect to carbon intensity. Based on the Stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT) model, we use Chinese industrial sector data from 2005 to 2016 to investigate how AI affects carbon intensity. The empirical results show that AI, as measured separately by the adoption of robotics by industry and the number of academic AI-related papers, significantly reduces carbon intensity. The results remain robust after addressing endogenous issues. We find that there are both stages and industrial heterogeneity in the effects of AI on carbon intensity. AI had a more decrease effect on carbon intensity during the 12th Five-Year Plan than the 11th. Compared with capital-intensive industries, AI tends to have a more decrease effect on carbon intensity in the labor-intensive and tech-intensive industries. To enlarge the effects of AI on reducing carbon intensity, the government should promote the development and application of AI and implement differentiated policies in line with the industry characteristics.</p>
economics,management,operations research & management science
What problem does this paper attempt to address?