AI and Higher Education: What Changes Now, What Changes Next
Across eight centuries, society has seen four big shifts: commerce, industry, the web, and now AI. Each one arrived faster than the last. This one is moving the fastest of all.
Generative AI hit mainstream use in under two years. Hundreds of millions now ask systems like ChatGPT for answers, drafts, code, and critique. On many campuses, a clear majority of students already use these tools for study, projects, and job prep.
What this actually is
Generative AI creates text, images, audio, and video from massive training sets. It can summarize, answer, classify, and draft on command. It's built on supervised learning and patterns derived from labeled examples, and it's getting easier to access in every tool your students touch.
The academic community often splits into two camps: alarmists who expect severe disruption, and non-alarmists who see big change but also clear upside. Both agree on one thing: ignoring this wave isn't an option.
Who feels the pressure first
Hands-on jobs with direct physical work are less affected in the near term. Routine cognitive work is hit first.
- Less affected, for now: barbers, EMTs, plumbers, orthodontists, painters, midwives, firefighters, landscapers, acupuncturists.
- Heavily affected: entry-level programmers, researchers and analysts, bookkeepers, lawyers' support staff, clerks and admins, translators, journalists and editors, PR, sales and call centers, warehouse and factory roles, quick-service restaurants, and drivers as autonomy expands.
What university leaders are saying
Boaz Ganor (Reichman University): Act now or live with a permanent skills gap
Ganor classifies himself as an alarmist by choice: better to be early than irrelevant. He argues universities must change three things at once: what they teach, how they teach, and the structure of learning itself.
Static degrees won't hold up against fast job cycles. Reichman is building multidisciplinary "cubes" of study that combine domain depth, humanities, teamwork, and computing. Four foundation courses are planned for all students: modern civics and public challenges, AI literacy and ethics, critical media and misinformation, and practical skills such as leadership and initiative.
David Passig (Bar-Ilan University): Fewer jobs, new models of learning
Passig forecasts a smaller share of the population in full-time work as automation grows. Some roles in hospitality and food service will move to robotics, while cities may expand short-format credentials people can complete from home.
He notes pilots for basic income show mixed results: after the initial relief, many people seek purpose and return to study. Expect confusion in standards before new norms settle.
Michal Feldman (Tel Aviv University): Keep the core - critical and algorithmic thinking
Feldman is measured, not alarmist. A CS degree still matters for structured thinking and research paths, but faculty must teach students to critique machine answers, not just consume them.
AI already removes the "productive struggle" from problem sets and writes essays on demand. That's a bigger issue in the social sciences and humanities, where intent and authorship are central. Human interaction still matters: people want meaning, not only a paycheck.
Daniel Chamovitz (Ben-Gurion University): Build AI resilience and double down on the campus
Chamovitz expects disruption but not the death of campus life. Zoom didn't replace classrooms; AI won't replace community. He's seeing renewed interest in philosophy and the humanities as students look for meaning.
His call: don't ban AI outright or outsource teaching to it. Use AI to strengthen what makes universities distinct - coaching judgment, process, and collaboration. Grade how students reached conclusions, not just the final text. Move past rote recall to genuine understanding.
Danny Raz (Technion): Update CS, rethink assessment, teach the AI dialogue
Industry leaders tell Raz the need for basic coders will drop, while demand for people who understand systems end-to-end will grow. Some will build core AI; many more will apply it inside fields like pharma and networks.
Technion now teaches how generative systems work to all CS students. Expect AI to help with grading and tutoring, but the priority is teaching students how to compare outputs across engines, ask better questions, and decide what's most correct.
Policy and national moves
Israel's Nagel Committee proposed a national AI strategy: a central directorate, a state supercomputer, and investment in talent pipelines. Critics flagged weak cost-benefit detail and limited industry representation. Creating a formal "AI ministry" was floated, but many argue universities and employers should lead with faster execution and fewer political bottlenecks.
A practical playbook for deans and department heads
- Set an AI-use policy that permits assistance with disclosure. Require students to note where and how AI helped, and to verify sources.
- Redesign assessment: in-class work, oral defenses, version histories, live data labs, and capstones where process and critique matter.
- Teach AI literacy across all majors: prompt strategy, bias detection, factual checks, and limitations. Make students argue against an AI answer before accepting it.
- Refresh programs on a 12-18 month cycle. Add short, stackable modules that reflect current tools without throwing out core theory.
- Push multidisciplinary fluency: pair law with data, biology with computation, design with analytics, economics with game theory.
- Invest in faculty development. Provide co-pilot tools and training. Run monthly clinics where instructors bring assignments and adapt them for AI-aware contexts.
- Upgrade academic integrity practices: require drafts, citations with links, and reflection memos; use small-group vivas to confirm authorship.
- Use AI where it helps: formative feedback, coding support, and simulations. Keep summative judgment with faculty.
- Center the campus experience: mentorship, peer studios, maker spaces, and real clients. Learning is social - protect that advantage.
- Mind data, copyright, and privacy. Prefer tools with strong safeguards and clear terms. Adopt departmental guidelines aligned with public standards such as the OECD AI Principles and UNESCO guidance on AI in education.
A 90-180 day rollout
- Days 1-30: Publish AI policy; run faculty workshops; pilot AI-assisted feedback in two courses per department.
- Days 31-90: Revise two core assessments per program; add an AI literacy module to first-year seminars; start student clinics on fact-checking and source tracing.
- Days 91-180: Launch cross-faculty project studios; formalize micro-credentials; set up an advisory council with employers to review updates each semester.
Key risks to manage
- Accuracy and fabrication: require source links and verification logs.
- Equity: provide campus access to approved tools so advantages aren't limited to those who can pay.
- Over-automation: use AI to assist, not to decide. Keep humans in final grading and advising.
- Faculty load: trade repetitive grading for higher-value mentoring; use assistants and AI for formative feedback only.
Where to upskill fast
If your team needs a quick way to scan current tools and credentials, see curated options by job role and certification tracks.
The takeaway for higher education
AI will squeeze entry-level cognitive work and raise the bar for judgment, synthesis, and collaboration. Degrees still matter, but process beats product, and community beats solo study.
Universities that adapt their content, methods, and structure will keep graduates employable and campuses essential. Those that wait for clarity will find the gap widening - and hard to close.
Your membership also unlocks:
