Can artificial intelligence curb greenwashing? Firm-level evidence based on large language model

Amid growing scrutiny of corporate environmental disclosures, concerns have intensified regarding the prevalence of greenwashing. Although the rapid advancement of artificial intelligence (AI) has drawn increasing attention for its transformative potential in corporate governance, its implications for environmental disclosure have only begun to receive scholarly attention and warrant further investigation. This paper investigates the impact of artificial intelligence adoption on corporate greenwashing using a panel dataset of Chinese A-share listed firms from 2011 to 2022. Leveraging a novel AI adoption index derived from a fine-tuned large language model (LLM), we conduct empirical tests to assess the relationship between AI use and firms’ greenwashing strategies. Our findings reveal that AI adoption significantly reduces the incidence of greenwashing, which remains robust across multiple validation checks. Decomposition analysis across different technological categories shows that planning and decision systems constitute the most influential strand of AI in curbing greenwashing. Mechanism analysis indicates that this effect operates through enhanced operational efficiency, improved human capital structure, and increased green innovation. Additional heterogeneity analysis across subsamples reveals that the deterrent impact exhibits greater intensity in firms characterized by non-state-owned firms, polluting sectors, and technology-intensive enterprises. By highlighting the governance potential of AI in promoting credible environmental disclosure, this study provides new empirical evidence on the intersection of digital transformation and corporate sustainability.