人工智能与企业ESG表现——基于国家新一代人工智能创新发展试验区的政策效应

Artificial Intelligence and Corporate ESG Performance—Based on the Policy Effects of the Construction of New Generation of National AI Innovative Development Pilot Zones

摘要: 人工智能作为引领新一轮科技革命的战略性技术，对企业可持续发展与ESG表现具有深远影响。本文以国家新一代人工智能创新发展试验区的批次设立为准自然实验，构建多期双重差分模型，基于2009~2024年沪深A股企业数据，系统考察人工智能创新政策对企业ESG表现的因果效应。研究发现：(1) 试验区政策显著提升辖区企业ESG表现，经多项稳健性检验后结论依然成立；(2) 机制检验表明，数字技术创新在政策与ESG表现之间发挥正向中介作用，媒体关注度则发挥不一致中介作用，正式制度供给对非正式舆论监督产生部分替代效应；(3) 融资约束对政策效应具有显著负向调节作用，行业集中度则发挥显著正向调节作用；(4) 政策效应在城市行政等级、数字关注度与市场化程度三个维度上存在显著异质性，制度环境相对完善地区的政策促进效果更为突出。本文为人工智能政策与企业ESG关系研究提供了新的经验证据，对优化人工智能创新政策设计与差异化推进ESG治理具有重要政策启示。

Abstract: As a strategic technology spearheading the new wave of technological revolution, artificial intelligence has a profound impact on corporate sustainability and ESG performance. This paper utilises the phased establishment of the National New Generation Artificial Intelligence Innovation and Development Pilot Zones as a quasi-natural experiment. By constructing a multi-period difference-in-differences model and drawing on data from A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2009 to 2024, we systematically examine the causal effects of artificial intelligence innovation policies on corporate ESG performance. The findings reveal that: (1) policies in the pilot zones significantly enhance the ESG performance of enterprises within their jurisdictions, a conclusion that remains valid following multiple robustness tests; (2) mechanism tests indicate that digital technology innovation mediates positively between policy and ESG performance, whilst media attention mediates inconsistently; formal institutional provision exerts a partial substitution effect on informal public oversight; (3) Financing constraints exert a significant negative moderating effect on policy outcomes, whilst industry concentration exerts a significant positive moderating effect; (4) Policy effects exhibit significant heterogeneity across three dimensions: urban administrative level, digital attention, and degree of marketisation, with policy promotion effects being more pronounced in regions with relatively sound institutional environments. This paper provides new empirical evidence regarding the relationship between AI policies and corporate ESG, offering important policy implications for optimising the design of AI innovation policies and promoting differentiated ESG governance.