From Fragmented Adoption to Systematic Transformation: Curriculum Analysis and an Institutional Roadmap for AI Convergence Education at a Korean University

Higher education institutions are under mounting pressure to transform their curricula in response to artificial intelligence (AI), yet many universities—particularly teaching-focused private institutions facing demographic decline—struggle to move beyond fragmented, course-level adoption. This study examines how a mid-sized private university in South Korea diagnosed and redesigned its major curricula for AI convergence education, framing this transformation as an institutional survival and quality strategy amid a shrinking college-age population. The study combined four methods: analysis of national AI education policies and of domestic and international university cases; a complete census of all undergraduate major curricula (67 program units; 1646 courses across the 55 non-AI-specialist majors) using a four-type typology of AI-linkable courses; a needs survey of department heads and program chairs; and rapid prototyping of a staged curriculum framework. The curriculum analysis indicated an articulation gap: explicitly AI-articulated courses accounted for only 1.5% of the curriculum (25 courses in 13 of 55 majors), whereas courses provisionally identified as plausible candidates for AI integration were nearly eight times as numerous (193 courses, 11.7%) and present in 84% of majors, but were not connected to curriculum documentation, learning outcomes, or progression pathways; whether these courses already contain AI-related practice requires syllabus-level verification. Two-thirds of responding program leaders (66.7%) endorsed AI convergence education while prioritizing concerns about infrastructure, institutional support, and unverified learning effects. Based on these findings, the study proposes the CROSS AX competency model—crossing disciplinary expertise with AI application capability—operationalized through a three-stage curriculum structure (AX-Foundation, AX-Bridge, AX-Field Lab) and a five-phase institutional roadmap moving from shared infrastructure toward full curricular transformation with quality management; both were expert-reviewed and iteratively refined over two review rounds. The study offers a diagnostic typology and roadmap that may be adapted by universities seeking to institutionalize AI-driven curricular transformation rather than accumulate isolated initiatives.