Abstract
Concerns surrounding trust, privacy, and security in the use of artificial intelligence (AI) in education are intensifying as AI tools become embedded in teaching, assessment, and institutional processes (Ayre et al., 2025; Lyu et al., 2025). These challenges are particularly acute in contexts where AI adoption advances more rapidly than formal governance mechanisms, contributing to fragmented and reactive institutional responses (Kumar et al., 2025). Although various AI governance frameworks have been proposed, institutional capacity is frequently cited yet rarely empirically specified, particularly across comparative geographical contexts.
This study examines the research question: What institution-wide strategies exist in universities to support and govern the integration of AI in pedagogy? A qualitative design combines reflexive thematic analysis with fuzzy-set qualitative comparative analysis (FSQCA) to identify both patterns and configurational pathways of institutional capacity. Data are collected from senior university practitioners involved in AI strategy and governance across the Global North and Global South.
Anchored in the structural dimension of social capital theory, the study conceptualises institutional capacity through formal governance structures, policy arrangements, and organisational networks. By identifying multiple pathways to AI governance, the findings contribute empirically grounded insights for strengthening responsible AI integration in higher education across uneven regulatory environments.