Configuration Paths of AI-Driven Urban Green Innovation: A Dynamic Qualitative Comparative Analysis Based on the TOE Framework

Artificial intelligence (AI) has increasingly been recognized as an important driver of green innovation, yet how cities leverage AI capabilities to achieve green innovation outcomes under different configurations of conditions remains insufficiently understood. This study develops a configurational framework based on the technology–organization–environment (TOE) perspective and applies dynamic qualitative comparative analysis (QCA) to panel data from 35 Chinese cities, comprising municipalities directly under the central government, provincial capitals, and sub-provincial cities, during 2019–2023. The results identify three configurations associated with high urban green innovation. AI Technology Development and AI Institutional Support consistently emerge as core conditions across all configurations, while AI Organizational Scale is identified as a recurring peripheral condition. The findings further reveal causal asymmetry between high and non-high levels of urban green innovation, showing that individual AI-related conditions alone are insufficient to achieve high green innovation outcomes without complementary conditions. Dynamic analysis indicates that the overall configurational structures remain relatively stable over time, although the consistency of individual configurations varies as AI applications and policy environments evolve. This study advances research on AI and green innovation by demonstrating that AI-driven urban green innovation depends on the interplay among multiple conditions and involves both stable foundational factors and evolving complementary conditions.