The role of generative AI in enhancing corporate ESG performance: evidence from China
TL;DR
The empirical results reveal that GAI adoption is positively associated with ESG performance, whereas DAI does not exhibit a similar relationship and offers valuable insights for policymakers and regulators into the distinct roles of generative and discriminative AI in shaping corporate ESG outcomes.
Abstract
Although corporate digital transformation has attracted increasing attention, limited research has examined how generative artificial intelligence (GAI) affects firms’ environmental, social, and governance (ESG) performance. Unlike discriminative AI (DAI), GAI is more closely associated with creativity, real-time feedback, and continuous interaction. To address this gap, we construct a firm-level GAI index using machine learning-based textual analysis and investigate its association with ESG performance among Chinese listed companies. Our empirical results reveal that GAI adoption is positively associated with ESG performance, whereas DAI does not exhibit a similar relationship. We further identify three mechanisms through which GAI exerts its influence: creativity stimulation, enhanced customer engagement, and improved operational risk management. Furthermore, the positive association between GAI and ESG performance is stronger among firms with higher intelligent investment, greater CEO digital literacy, and stronger internal controls, and is more pronounced in state-owned enterprises (SOEs) and firms operating in environmentally non-sensitive industries. These findings offer valuable insights for policymakers and regulators into the distinct roles of generative and discriminative AI in shaping corporate ESG outcomes.