Green innovation performance: dual impact of inventors’ knowledge characteristics and AI networks

Purpose Improving corporate green innovation performance is critical to sustainable development. Existing studies have paid limited attention to inventors’ knowledge structures and their underlying mechanisms in digital network contexts. Accordingly, this study aims to examine the effects of inventors’ knowledge diversity and knowledge uniqueness on green innovation performance and further explore the moderating role of artificial intelligence (AI) technology networks. Design/methodology/approach Grounded in the knowledge-based view and search-and-recombination theory, this study investigates listed firms in China’s strategic emerging industries. The study uses an inventor–knowledge two-mode network to capture inventors’ knowledge characteristics, incorporates the structural features of AI technology networks and tests the hypotheses using fixed-effects models. Findings The results reveal an inverted U-shaped relationship between knowledge diversity and green innovation performance, whereas knowledge uniqueness has a significant negative effect. AI technology networks further reshape the marginal effects of different knowledge characteristics, suggesting that knowledge value is strongly contingent on network embeddedness. Originality/value This study extends the knowledge-based view to the analysis of inventors’ knowledge characteristics. It explains the link between inventors’ knowledge characteristics and green innovation performance from a knowledge reconfiguration perspective. The study also conceptualizes AI technology networks as a digital knowledge architecture, thereby deepening understanding of green knowledge creation mechanisms in the digital context.