The smarter the cleaner: Evaluating the impact of artificial intelligence on haze pollution

This paper innovatively evaluates the level of artificial intelligence (AI) by counting AI firms and employs a two-way fixed-effects model to analyze the impact of AI on haze pollution (HP), using panel data of 264 Chinese cities from 2006 and 2021. It is found that AI is an effective means to reduce urban HP, and that for every 1 % increase in the number of urban AI firms, PM 2.5 is reduced by about 2.22 %. Regarding mechanisms, AI mainly reduces urban HP through two ways: stimulating technology innovation and promoting industrial structure upgrading. Heterogeneity tests show that the haze reduction effect of AI varies due to urban HP level, urban location and urban characteristics, and is more pronounced in cities with high HP, as well as in western cities, northern cities, cities of high administrative levels, and resource-based cities. Our findings provide new insights for policy makers to better utilize AI technology to reduce HP. • The impact of artificial intelligence (AI) on haze pollution (HP) is examined. • AI helps reduce urban HP, and that for every 1 % increase in the number of urban AI firms, PM 2.5 is reduced by about 2.22 %. • AI reduces urban HP through two ways: stimulating green technology innovation and promoting industrial structure upgrading. • The haze reduction effect of AI varies due to urban HP level, urban location and urban characteristics.