Beyond technology: Institutional foundations for AI-driven personalised learning in higher education

Purpose Successful AI-driven personalized learning depends as much on institutional conditions as it does on technology. Drawing on a sociotechnical systems perspective, this viewpoint explores the curriculum, governance and human capabilities needed for responsible and effective AI implementation in higher education. Design/methodology/approach The article draws on findings from two complementary empirical studies on students’ motivation and staff perspectives, alongside broader research on personalized learning and intelligent tutoring, to examine AI-enabled personalized learning through a sociotechnical systems perspective. Findings AI can build learner profiles and adapt content, pacing and feedback at scale, but its effectiveness depends on data quality and stops at the classroom door: socioeconomic status, family circumstances, and other off-platform pressures stay invisible to AI even though they are established drivers of attrition. Confidence, motivation and belonging stay central throughout, shaped by curriculum and educators as much as by any platform. Research limitations/implications As a conceptual synthesis rather than an empirical study, the conclusions require validation across different institutional contexts. Practical implications Universities should assess curriculum fit, strengthen governance, equitable access and AI literacy, and keep educators central when determining where and how AI-enabled personalization should be scaled. Originality/value This viewpoint reframes AI-enabled personalized learning as a discipline-sensitive institutional readiness problem, highlighting curriculum fit, AI literacy, algorithmic bias, equitable access, and the continuing importance of human care and judgment.