Conceptualising Human–AI Collaboration Competency for hybrid intelligence: towards learner agency through a metacognitive perspective

The growing integration of artificial intelligence (AI), especially generative AI and emerging agentic AI systems, in higher education raises urgent questions about curriculum design and the competencies required for meaningful human-AI collaboration. Although existing frameworks address technical and ethical dimensions, they neither clearly specify collaboration competencies within graduate profiles nor give sufficient attention to learner agency in AI-integrated learning contexts. This study proposes the Human-AI Collaboration Competency (HACC) framework, a holistic model that conceptualises human-AI collaboration as a dynamic configuration of disciplinary expertise, AI competency, and metacognitive competency. Rather than treating AI skills as discrete additions, HACC positions hybrid competency as central to students’ capacity to engage with AI critically, strategically, and responsibly. To examine its conceptual and practical relevance, we conducted semi-structured interviews with 24 educators (N = 24) across disciplines and role levels, generating 1,650 analysis units. Epistemic Network Analysis (ENA) was used to model structural connections among HACC dimensions and compare conceptual patterns across groups. The findings provide preliminary support for the framework and illuminate how related competencies are perceived and prioritised in practice. The study offers a basis for rethinking curriculum design in higher education that foregrounds disciplinary grounding and metacognitive competency as central to sustaining learner agency.