Can Artificial Intelligence Adoption Empower Digital–Green Synergy in Manufacturing Firms? Evidence from Large Language Models

Abstract

Under the dual pressures of resource constraints and emission regulations facing global manufacturing, digital–green synergy transformation has become crucial to achieving sustainable economic development. However, the existing literature still lacks micro-level evidence on how artificial intelligence (AI) adoption drives digital–green synergy (DGS). This paper takes manufacturing listed companies from 2011 to 2022 as the research sample. Based on the BERT large language model, an index of enterprise AI adoption is constructed, and a two-way fixed effects model is used to empirically test the impact of AI adoption on digital–green synergy in manufacturing firms. The results show that AI adoption significantly enhances DGS in manufacturing firms, with stronger impacts in high-tech, non-heavy-polluting, and non-state-owned enterprises. The mechanism analysis indicates that AI adoption mainly promotes DGS in the manufacturing industry through three paths: expanding knowledge breadth, optimizing resource allocation, and breaking through organizational routines. Further analysis indicates that the positive effect of AI adoption is reinforced by executives’ environmental backgrounds, while climate policy uncertainty exerts a dampening influence. Economic consequence tests confirm that improved DGS simultaneously enhances corporate social responsibility performance, fosters new-quality productivity, and strengthens supply chain resilience.