Quantifying the spillover effects of AI on green innovation and firm-level productivity: a stata GMM estimation approach

Purpose While macro-level studies exist, firm-level evidence on the dual role of AI in driving both productivity and green innovation particularly in South Asia is scarce. The purpose of this study is to quantify the spillover effects of AI on green innovation and firm-level productivity using longitudinal firm-level panel data. Design/methodology/approach The research is founded on the application of the longitudinal firm-level panel data which was recorded in plausible databases including World Bank key Fundamentals Enterprise Survey and Pakistan Bureau of statistics and reports of the state bank of Pakistan. It applies advanced methods of estimating dynamic panel data Proximate association between AI adoption, green innovation and productivity It applies both Difference Generalized Method of Moments (GMM) and System GMM. Inequality in the diffusion of AI on an industry level is also analyzed to ensure that the industry receives appropriate policy interventions. Findings Results suggest that AI plays a powerful role in both green innovation and productivity of a firm. A two-way linkage is identified: productivity advances through AI and this advances green innovation whereas green innovation slightly advances productivity. AI adoption is more in manufacturing and energy industries than in service activities and agriculture. The strength checks support statistically the validity of these results which reinforces the duality of AI in enhancing the economic performance and outcomes of sustainability in Pakistani firms. Originality/value This paper adds novel firm-level evidence from South Asia, where such studies remain limited. The results provide feasible information to policymakers and business leaders to transform digital transformation-sustainable development objectives in emerging economies.