Аңдатпа
Artificial Intelligence (AI) is increasingly part of higher education's institutional infrastructure, extending beyond teaching and learning to governance, administration, and student support. This study examines how AI Governance Readiness (AIGR) influences Administrative Efficiency through AI (AEAI), AI Fluency Orientation (AIFO), and Student Retention Intentions (SRI), while assessing the mediating roles of AEAI and AIFO and the moderating role of Equity Perception (EPS) in Kazakhstan's higher education sector. Drawing on Dynamic Capability Theory and Sociotechnical Systems Theory, the study develops and empirically tests an institutional capability model using 935 valid responses from students, faculty, academic administrators, and digital administration staff across public and private universities. The scholars analyzed the data using Partial Least Squares Structural Equation Modeling (PLS-SEM) in SmartPLS 4.0, supported by RStudio. The measurement model demonstrated satisfactory reliability, convergent validity, and discriminant validity, while the structural model showed adequate explanatory and predictive performance. We assessed common method bias through procedural safeguards and full-collinearity diagnostics, and identified no substantive concerns. The findings show that AIGR is positively associated with AEAI and AIFO, while both constructs positively influence SRI. AEAI and AIFO also significantly mediate the relationship between AIGR and SRI. However, EPS does not significantly moderate the relationships between AI-enabled institutional capabilities and retention intentions. The findings position AI governance as a strategic institutional capability through which universities can improve administrative responsiveness, develop AI-related competencies, and support student continuity. The study contributes empirical evidence from Kazakhstan to the growing literature on responsible AI governance and digital transformation in higher education.