Against the backdrop of global digital transformation and the rapid diffusion of intelligent technologies, artificial intelligence industrial agglomeration (AIIA) has become an important force reshaping regional innovation (RI) patterns. Yet systematic evidence on how emerging industrial agglomeration affects RI remains limited. Using panel data for 287 prefecture-level cities in China from 2011 to 2023, this study combines spatial analysis and econometric models to examine the impact of AIIA on RI and its mechanisms. The results show that RI level increased steadily, with a spatial pattern of coastal concentration and diffusion toward key inland nodes. AIIA also rose markedly, with high-level agglomeration areas concentrated in the eastern coastal region and gradually expanding outward. AIIA significantly promotes RI, and the result remains robust after multiple robustness checks. The effect is stronger in eastern regions, cities outside urban agglomerations, ordinary prefecture-level cities, and non-resource-based cities. Along the industrial chain, agglomeration across the upstream foundational, midstream technological, and downstream application layers promotes RI level, with stronger effects in the upstream and downstream layers. Mechanism tests show that human capital and digital economy development act as positive transmission channels. Government action and urban green space strengthen the innovation effect of AIIA, whereas innovation network embeddedness weakens its marginal effect. Threshold and spatial econometric results further indicate that AIIA has a critical-scale feature and significant positive spatial spillover effects. This study deepens the understanding of AIIA's role in shaping RI level and provides empirical evidence for optimizing artificial intelligence industrial layouts and enhancing regional innovation capacity.