AI Is Disrupting Higher Education: Are Universities Ready?

Adopt a binary AI policy: fully invigilated or fully AI-enabled. Boris Walbaum argues that the "grey area" on AI assessment is the single greatest risk to institutional credibility right now. Universities that allow some AI use while formally prohibiting it create an integrity gap that erodes trust with students and accreditors alike. The practical move is a clear, enforced split: either fully invigilated exams or assessments where AI use is explicitly permitted and itself assessed. There's no defensible middle ground.

Replace MCQ assessments with conversational AI testing. Multiple-choice exams measure recall but tell students almost nothing about gaps in their understanding. Walbaum points to conversational AI assessments as a concrete replacement: they probe reasoning, surface misunderstandings, and give students richer, more personal feedback than any automated scoring rubric can. Institutions reviewing their assessment strategy should pilot this format in courses where analytical thinking matters more than memorization.

Structure AI into the curriculum as the practice wall, not the classroom. Forward College uses AI, as Walbaum calls it, a "tennis practice wall": available for solo study, repetition, and preparation, but not during live class sessions. The classroom stays reserved for human interaction, debate, and the social reinforcement that actually rebuilds motivation to study further. Institutions rethinking blended learning can use this split as a practical design principle, not a vague aspiration.

Use AI as a student-persona simulator before courses launch. Before delivering learning materials, Walbaum suggests running them through AI with a student persona to identify confusion, gaps, and pacing issues. This is a low-cost quality check that most institutions aren't doing yet. It's most useful for newly designed courses or modules that haven't been tested with a real cohort.

Build curriculum around "deep skills," not just cognitive content. Walbaum's framework breaks student capability into four areas: cognitive, social, emotional, and practical. The research he cites from Cambridge shows that AI tools are already lifting the top 5% of performers but reducing motivation and confidence for a much larger share of students. A curriculum review that only asks "what do students know?" will miss the skills that hold students back and the ones that'll matter most when AI handles more of the cognitive load.

Address faculty resistance as an identity issue before rolling out AI tools. Walbaum's advice to hesitant higher ed leaders is to examine the emotional component of resistance first. For many educators, adopting AI isn't just a workflow change; it threatens their professional identity and sense of expertise. Deployment plans that skip this step and go straight to training programs tend to stall. The more practical approach is a structured conversation with faculty about what their value actually is, separate from the tools they've always used.

Reposition institutional excellence around learning communities, not research rankings. Walbaum's view is that research prestige rankings are backward-looking metrics that don't reflect what students are buying or what employers need. In the 5-year horizon he describes, the institutions that'll define excellence are those built around how well they form learning communities, not citation counts. For marketing and enrollment teams, this is a framing shift worth testing in messaging: prospective students who can't afford to bet on prestige alone are already asking different questions.