Exploring the Themes of Chinese Artificial Intelligence Policy: An <scp>LDA</scp> Topic Modeling Approach

ABSTRACT As a representative of next‐generation artificial intelligence, generative AI is profoundly transforming contemporary societal structures. As a pivotal player, China serves as both a primary application market and a key innovator in AI technology, with its developmental trajectory significantly shaped by national policy frameworks. This study employs Latent Dirichlet Allocation (LDA) topic modeling to systematically analyze 78 valid and currently implemented AI policy documents in China. The research aims to identify core focus areas in China's current AI policy landscape and provide insights for sustainable development of AI. Analytical results highlight seven key policy themes: (1) technological innovation and industrial integration, (2) social governance and mechanism evaluation, (3) model training and disciplinary methodologies, (4) software algorithms and data security, (5) pilot zone construction and innovation development, (6) infrastructure and intelligent service systems, and (7) AI research project implementation. Based on these findings, the study concludes with targeted policy recommendations.