Human AI Collaborative Subject Design and Assessment Workflow

The rapid adoption of generative artificial intelligence in higher education has increased instructional efficiency but has also raised concerns regarding academic governance, pedagogical control, and ethical accountability. While existing studies largely focus on tool adoption and user perceptions, there remains limited empirical evidence on how instructors systematically interact with generative artificial intelligence during authentic course design, teaching, and assessment processes. This study addresses this gap by proposing and empirically examining the Human-AI Collaborative Subject Design and Assessment Workflow, a human-centered instructional framework that structures collaboration between instructors and generative artificial intelligence across the full instructional lifecycle. Grounded in Design Science Research, the framework was designed, enacted, and evaluated through full-semester implementation in an undergraduate university course. Empirical evidence was derived from authentic instructional artifacts, including syllabi, lecture materials, laboratory activities, and assessments, as well as documented human–AI interaction processes. The results demonstrate that generative artificial intelligence can effectively support drafting and structuring tasks while human instructors retain exclusive control over pedagogical validation, laboratory design, assessment finalization, and grading. The findings confirm the feasibility of workflow-based, human-in-the-loop integration under real teaching constraints, showing that instructional scalability and efficiency can be achieved without compromising academic authority, ethical governance, or pedagogical quality. This study contributes a replicable, instructor-enacted model for responsible generative artificial intelligence adoption and advances process-level governance approaches for higher education teaching practice.