From Clarity to Conviction: Instrumental Limits and Integration Pathways for Generative Artificial Intelligence in University Ideological and Political Education

This qualitative study examines how generative artificial intelligence is being integrated into university ideological and political education (IPE) in China and delineates the conditions under which its instrumental rationality reaches its practical limits. We conducted semi-structured interviews with 17 instructors from five universities in Chongqing (45–120 minutes, in Chinese), audio-recorded, transcribed verbatim, and analyzed using reflexive thematic analysis (RTA). Sampling and stopping were guided by information power; we judged data adequacy when the developing patterns were sufficiently rich and useful for the research questions. NVivo 12 supported data management. We identified three themes: attenuation of affective and faith dimensions; content complexity and the limits of AI understanding; and insufficiency of high-quality, compliant training data. Building on these findings, we propose an integration framework that aligns classroom practice with platform support and institutional governance, and we formulate actionable recommendations for policymakers, universities, and instructors.