Efficiency Tool or Academic Threat? Mapping China’s AI Controversy in Higher Education with Mixed Methods

Global higher education is experiencing dual pressures from rapid rise of generative AI tools such as ChatGPT, which drive pedagogical innovation while triggering global concerns over academic integrity and skill degradation. This study addresses a key research gap by investigating how China’s distinct context, marked by government influence, market dynamics, and achievement pressure, influences conflicting understandings of AI in academia. Employing a mixed-methods approach combining Latent Dirichlet Allocation (LDA) topic modeling and Social Construction of Technology (SCOT) framework-guided content analysis, discourse dynamics among key social groups was mapped: producers, advocates, users, and bystanders. Thematic modeling identified five key areas of controversy: AI-enabled academic practices, ethical concerns, efficiency evaluations, policy debates, and the perceived inevitability of AI adoption. Content analysis revealed clear differences in perspectives among key groups: producers highlighted efficiency benefits in their communications; advocates stressed academic integrity risks and endorsed regulatory measures; users explained their reliance on AI as necessary due to academic pressures and curricular dissatisfaction; while bystanders expressed diverse but often practical views on AI’s broader societal impact. Crucially, our analysis indicates that China’s closure mechanism will likely be predominantly policy-driven, advocating tiered governance and pedagogical reforms to mitigate structural pressures fueling student dependency. This research contributes a non-Western perspective to SCOT theory, revealing how administrative-market dynamics shape technology negotiation. Furthermore, it proposes practical implementation pathways for adaptable AI governance frameworks applicable across diverse global higher education contexts.​