Generative Artificial Intelligence-Supported Project-Based Learning in Chinese Vocational Digital Media Education: A Teacher-Regulated Conceptual Framework

HRMARS - Generative artificial intelligence (GenAI) can help digital media students interpret tasks, locate information, develop ideas, produce prototypes, and resolve technical problems. However, it can also perform the thinking and production that students are expected to undertake themselves. For project-based learning, the central issue is therefore not whether artificial intelligence should be used, but where assistance is educationally appropriate, which responsibilities must remain with students, and how learning can be demonstrated. This conceptual paper integrates project-based learning, sociocultural scaffolding, cognitive load theory, self-regulated learning, and recent research on generative artificial intelligence in education. It proposes a teacher-regulated framework for Chinese vocational digital media education comprising five stages: problem framing, evidence inquiry, ideation and prototyping, production and collaboration, and assessment and reflection. Across these stages, teachers establish boundaries for artificial intelligence use, examine process evidence, and gradually fade support as students become more capable. The framework's novelty lies in integrating stage-aligned assistance, retained student responsibility, process evidence, and scaffold fading under teacher regulation. By treating GenAI use as an organisation of agency and responsibility rather than tool adoption, it contributes to social science research on GenAI-mediated education and offers five propositions for curriculum design and empirical testing without claiming verified effects.