Colleges and universities are revising cloud and data policies as AI workloads move beyond research into student services, admissions, advising and campus operations. IT leaders report that decisions about where AI models run are now driven by data-privacy rules, vendor contracts and institutional risk assessments rather than pure cost or performance. The shift is prompting new governance structures that tie procurement, legal counsel, and academic units together to manage compliance and liability. Industry observers say the change is widening the role of campus CIOs: routine cloud decisions now require alignment with institutional counsel, IRBs and privacy officers. Leaders are also weighing on‑premises options to keep personally identifiable student data under tighter control, while negotiating bespoke terms with hyperscalers for model access and data residency. Institutions with large research AI programs are piloting hybrid models to balance agility and oversight. The broader challenge is institutional absorption: countries and organizations that built strong governance and identity frameworks—like India’s Aadhaar-backed systems in government services—offer case studies of the nontechnical work required to scale AI. University leaders told their staffs that adopting AI at scale will require changes in incentives, staff training and procurement timelines, not just faster GPUs.