Building an independent knowledge system in China is a strategic proposition for accelerating the development of philosophy and social sciences with Chinese characteristics, which holds profound significance for enhancing China’s theoretical innovation and cultural soft power. As a representative of cutting-edge productive forces, artificial intelligence (AI), through its technological breakthroughs and independent innovation in academic research, not only provides methodological support for transforming the knowledge production paradigm in China’s philosophy and social sciences but also serves as a crucial focus for China to break Western technological monopolies and reduce excessive reliance on foreign academic pathways. Therefore, conducting research on the evaluation of China’s international academic discourse power in AI and establishing a comprehensive, diversified, and multi-level evaluation system will contribute to advancing the development of China’s philosophical and social sciences in terms of disciplinary, academic, and discursive systems. It will also facilitate the transformation and application of AI paper achievements and support the construction of China’s independent knowledge system. First, this paper focuses on Chinese AI papers as the research object. It conducts an in-depth analysis of the specific connotations and constituent elements of the evaluation dimensions for China’s international academic discourse power from four perspectives: academic influence, academic perception, academic communication, and academic leadership. Second, building on the dimensional analysis, the study draws on Foucault’s theory of discourse power, Fei Xiaotong’s theory of cultural self-awareness, and Braddock’s 7W model to explore the generative logic of international academic discourse power and clarify its underlying mechanisms. Third, by collecting multi-source heterogeneous data and integrating the strengths of diverse methods, the study employs the CRITIC method and the projection pursuit classification model to conduct both single-dimensional and comprehensive evaluations of the academic influence, perception, communication, and leadership of China’s AI discourse power. In the comprehensive evaluation, assessment is carried out separately for three major disciplines: computer science, chemistry, and environmental science. Finally, convolutional neural networks and correlation analysis are used to validate the evaluation results, robustly demonstrating the effectiveness of the proposed system in measuring China’s discourse power in AI in the international academic arena. This study aims to enrich and refine the theory and indicator system for evaluating China’s international academic discourse power in AI, thereby contributing to the construction of an independent knowledge system and enhancing China’s voice in global AI academic discourse.