Abstract: Understanding the intricate roles of artificial intelligence (AI) is crucial for enhancing environmental, social and governance (ESG) performance. For this purpose, this paper adopts a methodology encompassing full and sub sample approaches to delve into the intricate connection between AI and ESG. The empirical results underline that AI has both a facilitative and an inhibitory influence on ESG. The positive aspect indicates that, under certain circumstances, AI can potentially boost ESG performance. However, negative impacts, such as privacy violations, algorithmic bias, energy consumption, job losses, and regulatory risks, counteract this. In addition, when AI is relatively stable, ESG might improve due to policies, investor preferences, corporate awareness, and scrutiny. Thus, AI does not consistently enhance ESG performance. ESG, in turn, has a two-way impact on AI. While it boosts AI by fostering demand for safe, eco-friendly tech, increased ESG scrutiny and trade wars crowd out investment, harming AI funding and market confidence. Amid a new technological revolution, we formulate targeted policy recommendations designed to leverage the beneficial aspects of AI while minimising its detrimental impacts on ESG performance.

Keywords: Artificial intelligence; ESG performance; Time-dependent relation

JEL codes: C32 O33 Q56

DOI: ...