Artificial intelligence adoption and green collaborative innovation: knowledge-based channels and governance conditions
Abstract
Purpose This study aims to examine whether and how firm-level artificial intelligence (AI) adoption promotes green collaborative innovation. It focuses on two knowledge-based capability channels, namely, knowledge learning capability and green governance capability, and investigates the organizational and institutional conditions under which the innovation effect of AI varies. Design/methodology/approach This study uses panel data on Chinese A-share listed firms from 2012 to 2023. AI adoption is measured by per-capita machinery and equipment book value, and green collaborative innovation by jointly filed green patents. The analysis applies two-way fixed-effects models, robustness tests, instrumental-variable estimation and propensity-score matching. Findings AI adoption significantly promotes green collaborative innovation in both invention and utility model patents, and the results remain robust across multiple tests. It strengthens knowledge learning and governance capabilities, while internal green strategy and external media coverage amplify its effect. The effect is stronger among firms with broader patent knowledge bases, high-technology and heavily polluting firms, and those located in cities with better digital infrastructure. Originality/value This study shifts attention from AI as an efficiency-enhancing technology to its role in reshaping firm knowledge processes. It identifies knowledge learning capability and green governance capability as two pathways through which AI supports green collaborative innovation, and uses knowledge-distance governability to explain how strategic commitment, media scrutiny, knowledge endowments, industry characteristics and digital infrastructure shape AI’s innovation returns. Using knowledge-distance governability, this study explains how AI makes distant knowledge more visible, interpretable and applicable to green innovation.
