Artificial Intelligence Fueling Endogenous Innovation: Evidence on Global Value Chain Upgrading in Chinese Manufacturing Firms

This study investigates how artificial intelligence (AI)-driven endogenous innovation enables Chinese manufacturing firms to upgrade their positions in global value chains (GVCs). Based on survey data from 287 firms, we identify a core mechanism through which AI alleviates resource constraints by improving technical efficiency, supporting data-driven decision-making, and facilitating knowledge recombination. This mechanism helps firms overcome low-end lock-in and move toward higher value activities. Our analysis reveals two key findings that contrast with established views. First, the primary internal driver of innovation is organizational innovation culture rather than individual entrepreneurship, refining the traditional Schumpeterian paradigm's emphasis on the entrepreneur. Second, while absorptive capacity strengthens process and product upgrading, it does not support functional upgrading, revealing a disconnect between technological capability and governance power. The study contributes theoretically by clarifying the linkages among AI capabilities, endogenous innovation, and GVC upgrading. For managers, it underscores the importance of cultivating an innovation-oriented culture within the organization, while leveraging external market pressures and policy support to build a robust foundation in data, algorithms, and computing power. All findings are validated through structural equation modeling and robustness checks, providing reliable insights for both research and practice.