AUTHOR=Wang Jianming , Li Wei 

TITLE=Shift one's trouble to others: Does climate policy uncertainty promote enterprises' “pollution migration” in the context of artificial intelligence?

JOURNAL=Frontiers in Public Health

VOLUME=Volume 13 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2025.1578139

DOI=10.3389/fpubh.2025.1578139

ISSN=2296-2565

ABSTRACT=Against the backdrop of worsening global climate change, countries worldwide have implemented climate policies to reduce corporate pollution emissions and promote corporate social responsibility. However, regional differences in climatic conditions have intensified the uncertainty of climate policies during implementation, creating a critical research gap: the influence of climate policy uncertainty (CPU) on corporate pollution behavior remains underexplored, despite its theoretical value for enriching environmental policy and corporate behavior research and practical significance for guiding policy optimization. To address this gap, this study takes 3,702 listed enterprises across 31 provinces in China (2010–2022) as the research sample. It empirically examines the impact of CPU on enterprises' “pollution migration” behavior, with a focus on testing underlying mechanisms (e.g., financing constraints) and heterogeneous effects (e.g., by artificial intelligence [AI] adoption level, enterprise pollution intensity, and ownership type). The key findings are as follows: (1) CPU significantly exacerbates enterprises' pollution migration; (2) the mechanism test confirms that CPU increases enterprises' financing constraints, which in turn aggravates pollution transfer; (3) enterprises with higher AI adoption levels experience a weaker impact of CPU on pollution migration; and (4) heterogeneity analysis shows that CPU exerts a more pronounced effect on pollution migration among highly polluting enterprises and non-state-owned enterprises (NSOEs). This study validates the “pollution haven” hypothesis in the context of climate policy uncertainty, providing important references for both policymakers and enterprises. For governments, it is recommended to stabilize climate policy expectations, improve the green financial system, and support enterprises in AI application. For enterprises, proactive monitoring of policy trends and enhancement of AI application capabilities are essential to mitigate the adverse effects of CPU and achieve sustainable development.