A Reference Model For Integration of Large Language Models In Higher Education: A Study of Doctoral Students’ Perception of The Model’s Applicability and Relevance

The integration of Large Language Models (LLMs) in higher education is accelerating, yet existing frameworks rarely provide a cross-dimensional integrated approach that aligns technical, pedagogical, ethical, and organizational dimensions, resulting in fragmented adoption. This research addresses this gap by proposing a technology-agnostic, four-stage reference model that includes two innovative components: a maturity model (LLM-HE-MM) to assess institutional readiness and a specialized LLM Cluster to support coordinated human–AI collaboration. A focus group with doctoral students and lecturers was conducted for preliminary empirical validation of the proposed model. The findings indicate that the model is perceived as clear and useful, while also confirming concerns regarding the impacts of LLM use for data privacy, intellectual autonomy and critical thinking.