Higher Education

Higher Education — the integration of artificial intelligence into university teaching, learning, assessment, and administration. Higher education is the most-studied context in the knowledge base, with over 100 articles examining how AI transforms college-level instruction, institutional policy, and student experience. AI in higher education is both the dominant setting for AIED research and the site where its tensions are most visible — between generative AI's promise of scalable personalization and its risks to integrity, learning, Privacy, and equity. It encompasses both undergraduate and graduate study, including professional programs such as medicine and business.

Questions to Consider

• Higher education is where AI's tensions are most visible — scalable personalization versus risks to integrity, learning, privacy, and equity. Which of these tensions do you see playing out most concretely in your own institution?
• Large studies reveal gaps between institutional policy and what students actually do with AI day to day. Why do you think official policy so often diverges from real student practice?
• The page frames AI adoption as uneven — some institutions transform while others lag, shaped by organizational drivers and national policy. What factors do you think determine whether a university adapts or stalls?
• As AI makes personalized one-on-one support nearly free and universal, how does that shift the value and role of the university itself? What becomes scarcer and therefore more valuable?
• One finding notes a decline in study time among college students using AI. If students spend less time studying but report satisfaction, what does that suggest about what institutions should be measuring and guarding?
• Given the gap between faculty self-assessed and actual AI readiness, what would meaningful faculty development look like in your context — and who should be responsible for it?

Introduction

AI in higher education research spans every function of the university: from AI tutoring and automated grading to faculty development, academic integrity, institutional governance, and student support. The knowledge base's higher education articles cluster around several key themes — institutional transformation, student experience at scale, faculty and teaching, assessment and integrity, and the rapid shift in policy and practice. Across the review's 394 GenAI higher-education studies, participant focus was students in 352, instructors in 15, both in 26, and multiple stakeholder groups in 1 — so the evidence base for perceptions and teaching practice is overwhelmingly student-based (Lin et al. (2026)).

Institutional transformation

Institutional change frameworks analyze how universities adapt to AI — not just at the classroom level but across policy, governance, and organizational structure. The EPIQ-AI framework reframes faculty readiness as a sociotechnical alignment challenge involving epistemic, pedagogical, institutional, and quality domains. Rethinking universities in the AI era examines whether current institutional models can accommodate AI-driven education. Adoption is not uniform: organizational-driver research identifies what enables or blocks institutional uptake, and national policy analyses show how systemic context shapes university responses. A concrete institutional blueprint comes from Qin (2026), who documents Lingnan University's repositioning as a "Research-Intensive Liberal Arts Institution in the Digital Era," mandating GenAI literacy for all undergraduates and embedding digital literacy across the Common Core while developing a human-in-the-loop model that foregrounds ethical reasoning and critical judgment — arguing the AI-for-education shift is an intellectual transformation, not technocentric augmentation. Macro-level mapping of 22 higher-education AI-integration interventions (AlSheikh et al., 2026) reinforces that transformation remains more aspiration than reality: graded on the SAMR model, most interventions sit at Substitution or Augmentation, integration is concentrated in medicine, engineering, and computer science with little humanities presence, and evidence skews to North America and Asia with minimal Global South representation.

A global comparative index measures the discourse–structure gap. Chan (2026) coded the official websites of all 100 Times Higher Education 2026-ranked universities against a 20-indicator AI Institutionalization Index. The symbolic layer is saturated while the structural layer is thin: 100% have an AI task force and a master's AI degree, but only 12% have executive AI leadership, 5% require an introductory AI course for undergraduates, and 0% publish an AI accessibility policy despite 93% deploying AI campus services.

Institutionalization also proved independent of prestige (rank correlation r ≈ 0.04), with significant regional variation (ANOVA F = 5.86, p < .001; North America M = 0.63 vs. Europe 0.51, Asia 0.50) and a six-model typology in which only a small subset of mostly large US publics reach "deep institutionalization." The author flags the index as a first-generation, unvalidated, website-only measure of the elite top 100 — a comparative map, not a scorecard.

Which forms of GenAI use an institution has made legitimate is itself measurable. Sun, Dohn and Rehm (2026) classified 102 empirical higher-education studies by how epistemic work was distributed between student and system, and found Assistant (42.16%) and Enabler (42.16%) applications dominating while Generator, Reimaginator and Synthesizer together made up under a fifth of the corpus — not because the models cannot do the rest, but because the marginal forms distribute authorship in ways the university's individual-competence logic does not yet recognize. On their account the binding constraint on GenAI in higher education is institutional legitimacy rather than technical capability, which puts the question squarely in the same policy and assessment-redesign territory as integrity and Assessment reform.

Ranieri and Biagini (2026) tested the 2019 four-category higher-education taxonomy against 228 publications from 2019–2026 and found 58.4% of the 154 classifiable applications fall outside it, adding AI-enabled student support, content and instructional design, and institutional analytics as functional domains.

AI deployment outside the classroom has become a consumer-protection question, and the student-side inventory of it is thin. Student Defense's SHAPE AI report (2026) maps institutional AI use across admissions, recruitment and financial aid, student-success services (advising, counseling, benefits access, credit transfer), and education delivery — LMS-embedded features from Canvas, Blackboard and D2L Brightspace, institution-wide GenAI platforms, tutoring and feedback systems, detection and proctoring — and argues that adoption is outrunning protection. Its six cross-cutting concerns (whether students get what they pay for, the loss of learning communities, disclosure and transparency, data privacy, bias, and calibrated human oversight) are joined to a four-question procurement screen: what problem is being solved, whether governance capacity exists, whether the institution is being transparent, and whether the deployment creates value for students. The report's position is that states, accreditors and the federal Department of Education are not yet equipped to catch these risks, leaving the institution's own written policy as the operative protection, and it names the failure mode directly: a college marketing a faculty-led program while instruction or course design is outsourced to AI or third-party vendors is selling a different product under its own name.

Student experience at scale

Large-scale studies of authentic student AI use and GenAI availability and satisfaction document how students actually use AI — revealing gaps between institutional policy and everyday practice. Survey research captures how students navigate the AI landscape, and studies of study time, AI skills for graduates, and student perceptions of AI tools show that the student experience of AI is mixed — efficient but often shallower. Much of this university learning now happens online, where online teaching and learning shapes both AI's benefits (scalable personalization, always-on support) and its risks (academic integrity, cognitive offloading) for college students. An exploratory ML approach using SHAP analysis examined how students' perceptions and demographics relate to intended academic ChatGPT use, prioritizing interpretability (An Exploratory Machine Learning Approach to Understanding Determinants of Future ChatGPT Use in Higher Education). In art and design education, where generative AI challenges the value of human Creativity, a project-based learning model with digital storytelling at its core cultivated the emotional, cultural, and narrative capacities AI lacks — evaluated through a 15-week embedded case study with 426 Chinese undergraduates who translated local cultural heritage into Multimodal AI narratives (In the AI era: A project-based digital storytelling framework for art and design education). Assistive and inclusive uses of GenAI are also reshaping the student experience for disabled learners: Khlaif et al. (2026) — a qualitative case study of 21 visually impaired undergraduates across three Palestinian universities — found GenAI tailors pace, content, and delivery to individual profiles, simplifies complex texts, and converts content across modalities, giving learners a sense of autonomy and confidence in class participation while being viewed as complementing rather than replacing teachers.

A hierarchical regression on 303 undergraduates found perceived resources and teacher support explained 29% of AI-literacy variance, prior digital competence 3%, the psychological block 6% (total R² = 0.39); technology anxiety's association reversed to positive in the full model, only for critical evaluation and ethical competence (Chow et al. (2026)).

Beyond the classroom, AI reaches students through the institution itself — advising, outreach, referral, and the allocation of scarce support — which student support and success gathers as its own body of evidence. Its central caution about scale comes from a four-year randomized evaluation (N = 8,708) in which registration nudges moved a deadline by 34 percentage points while graduation, credits, and GPA did not move at all.

Graduate and professional education

The knowledge base is weighted heavily toward undergraduate study, and it houses graduate and professional preparation inside discipline-specific pages (medical and health-professions education, business education, teacher education) rather than in a separate graduate-education page. Two strands nonetheless bear specifically on master's, doctoral, and professional students: research training, and professional-degree programs.

Postgraduate research practice is the more distinctive of the two. Dai and Chan (2026) analyzed how 28 postgraduate research students across seven focus groups enacted AI Literacy in their GenAI-assisted research and found them to be calibrated users rather than passive adopters — matching tools to tasks by perceived stakes and disciplinary norms, retaining human oversight, and drawing their own boundaries between acceptable drafting help and substitution of their reasoning. Their central finding is a policy gap: institutional GenAI guidance addresses teaching, learning, and assessment but says almost nothing about the research process, so the authors propose researcher-oriented guidelines that scaffold each AI literacy dimension across the research workflow for researchers and their supervisors. Oh, Talton and Bui (2026) approach the same population from the institutional side: among 93 bioscience graduate students and postdoctoral trainees in a required research-ethics course, three simple pre-instruction behavioral signals informed lightweight intake profiling for adaptive AI Ethics instruction, while prior AI coursework predicted none of the five perception outcomes. AI is also reshaping the research team itself — analysis of 147,074 publications since 2020 associates AI-assisted writing with smaller, junior-leaner teams producing highly cited work, reversing the decades-long "Big Science" drift toward larger collaborations — and a single-investigator case study documents what sustained agentic support looks like inside one researcher's own workflow.

Professional-degree programs hold most of the graduate-level evidence, and it is organized by discipline. In health-professions education, a randomized controlled trial of 100 third-year medical students tested a multi-agent LLM system — simulated patient, Socratic tutor, turn-level evaluator — organized around explicit Scaffolding functions for clinical interview training, while the SCAN framework reframes trainee AI use as a problem of task distribution and real-time metacognitive classification rather than learner misuse. Master's-level preparation appears in Zhu et al. (2026), whose D–T–E model raised mathematics-education M.Ed. students' instructional-design competence (Cohen's d = 0.62) by having three LLMs act as "intelligent reviewers" in a design–feedback–reflection loop. How students conceptualize GenAI and their own professionalisation cuts across these programs, while workplace and continuing professional learning — adjacent to but distinct from graduate study — is treated under Adult Learners and Workplace Learning.

Faculty and teaching

Educational Development research examines how instructors adopt, resist, or adapt to AI. Teacher AI adoption studies identify confidence, support, and attitude as key predictors. AI-assisted discretionary feedback research explores whether AI increases instructor feedback quality and quantity, and ethics research weighs whether teachers should use AI for feedback at all. Staff-perspectives research and phenomenographic studies of teacher conceptions probe how educators understand their changing role. In large-class settings, AI can now shoulder a growing share of the feedback load: a semester-long field experiment in undergraduate macroeconomics tutorials (Geschwind et al., 2026) found that individual GPT-4 feedback sustained the highest participation across eight open-ended tasks and produced the largest content learning gains, positioning AI feedback as a scalable complement to lecturer feedback and a substitute for unreliable peer feedback. A complementary route for large undergrad STEM classes is a diagnostic teaching-assistant system: Arthur (Yin et al. 2026) uses a per-question ML backbone to give real-time, personalized feedback on Engineering Economics Calculated Formula Questions — a domain where handwritten solutions had previously blocked AI support — via a dialogue-based, question-bank web interface that balances feedback accuracy against collection efficiency. Farazouli et al. (2026) capture the emotional dimension of that change: 24 Swedish university teachers experienced GAI's emergence as alarming and overwhelming, reporting a "state of vulnerability" (low confidence, insecurity, fear of "not being ahead of students") and feeling "stuck" between utopian and dystopian discourses — while rethinking Assessment and re-evaluating their priorities toward critical thinking and ethical GAI use.

One hour of AI tutoring was statistically equivalent to live expert human tutoring on GRE learning gains across 2,469 sessions with 2,383 adults, and one model reached parity at 918× lower cost per percentage point (Northcutt et al. (2026)).

Assessment and integrity

Academic Integrity and AI assessment reform research grapple with how universities should redesign evaluation for an AI-capable student body. Detection-centered approaches are giving way to Authentic Assessment and process-based evaluation, including oral exams, assessment twins, and authentic assessment redesign. The shift reflects a deeper concern: how students and institutions define cheating with AI, and the quality of AI-generated assessment materials. Automation of grading itself is maturing for open-ended university work: Pecuchova, Benko & Drlik (2025) benchmarked eleven GenAI and sentence-embedding models on 1,885 responses to 24 software-engineering exam questions and found only GPTo1 reached almost-perfect agreement with two expert human graders (Fleiss' Kappa 0.82), with context-sensitive models outclassing reference-based ones — but its proprietary API costs led the authors to recommend hybrid human-in-the-loop deployment for resource-constrained institutions. At the professional-program level, Olvet et al. (2026) show that Large Language Models (LLMs) (GPT-4) scoring of pre-clerkship medical open-ended exam questions can reach substantial-to-almost-perfect agreement with faculty graders (weighted kappa up to 0.94) after iterative human rubric refinement, while the authors still recommend keeping humans in the loop to arbitrate residual discrepancies. Complementing this model-level evidence, Falahat, Das, Bhaumik & Thambi (2026) graded a 21-item university pharmacy exam with ChatGPT-5 and found near-perfect agreement with faculty on objective items (CCC 0.935–1.000) but unreliable agreement on short-answer and essay items, with rubric provision not consistently improving performance — reinforcing the case for hybrid, human-in-the-loop grading in university assessment.

A third route hardens the assignment itself: rendering prompts as perturbed images raised LLM assignment-completion failure from 7.4% to 55.6% and pushed Tesseract extraction error to 100.0%, reframing integrity as content design that complements detection rather than replacing it (Salaun et al., 2026).

Implications for higher-education instructors

• Design assessment for an AI-capable student body. Detection-centered integrity approaches are giving way to authentic and process-based evaluation (beyond detection, assessment reform) — redesign what you assess, not just how you police it.
• Stop treating an in-room ban as assurance. Smart glasses and AI-enabled hearing aids are dually transparent — no gaze drop or posture shift to detect — so an invigilated exam or oral can no longer purchase evidential confidence by physical separation (Corbin et al., 2026).
• Redesign the "what should students still learn by hand?" question. Just as in computing, decide which skills must be preserved (verification, judgment, process) and make those the assessed core, rather than assuming AI skills transfer automatically.
• Treat faculty readiness as sociotechnical, not just technical. EPIQ-AI frames readiness across epistemic, pedagogical, institutional, and quality domains — align your teaching practice with institutional governance, not just tool fluency.
• Fund AI literacy, not just tool confidence, for faculty. Liu et al. (2026) found that social and organizational support reached university faculty AI well-being only indirectly, through AI literacy (β = 0.713), so programs that build conceptual, ethical and pedagogical understanding do more for faculty than hands-on fluency alone.
• Address the policy-vs-practice gap. Large-scale studies (AI in the wild) show students use AI in ways institutional policy doesn't anticipate — align your expectations with real usage and teach AI Literacy explicitly.
• Write method-specific guidance, not one campus-wide AI policy. Interviewing 25 university language teachers, the study distilled 20 ethical tensions — four per online method — grouped into authenticity of learning (11), authority challenge (4) and collaborative dynamics (5), and argues that generic, method-agnostic policy cannot address the distinct pressures each online method places on integrity and collaboration (Baoyi & Khan (2026)).
• Use AI to raise feedback quality and quantity. AI-assisted feedback can increase the feedback instructors deliver; pair it with human judgment so it improves learning rather than merely automating.
• Engage with grading and assessment innovation deliberately. Mesny, Roberge-Maltais & Galy (2026) argue that Assessment and grading are among the most influential factors shaping learning in higher education, yet they remain under-researched in fields such as management education (only 58 articles over 20 years across four leading journals). Traditional, summative-heavy and norm-referenced approaches undermine deep learning, Well-Being, equity, and integrity in the generative AI era; the authors urge instructors to engage more actively and reciprocally with five innovative practices (authentic assessment, self- and peer-assessment, reassessment, standards-based grading, ungrading), noting implementation demands institutional and cultural support and incremental experimentation.
• Treat research training and supervision as an AI-literacy site in its own right. Postgraduate researchers enact AI Literacy as a situated practice rather than a static skill set, and institutional guidance largely stops at teaching and assessment. Guidance for thesis, dissertation, and publication work should scaffold verification, attribution, and disclosure concretely for each AI literacy dimension — and simple pre-instruction signals can target that support in research-ethics courses — instead of issuing binary rules.

Common Misconceptions

• "Institutional AI adoption has matured." The symbolic layer is saturated while the structural layer is thin: 100% have an AI task force and a master's AI degree, but only 12% have executive AI leadership and only 5% require a further institutional step. Announcements have outrun structures by an order of magnitude.
• "Prestige predicts institutional AI maturity." Institutionalization proved independent of prestige (rank correlation r ≈ 0.04), with significant regional variation (ANOVA F = 5.86, p < .001). Reputation is not a proxy for readiness, and geography predicts more than status.
• "How AI distributes epistemic work is uniform across institutions." Sun, Dohn and Rehm (2026) classified 102 empirical higher-education studies by how epistemic work was distributed, and the variation is the finding rather than a detail.

Connected Concepts

• Pedagogical Patterns — The level where most of the tested sequence evidence was generated
• Online Teaching and Learning — Online Teaching and Learning
• Generative AI — Generative AI technologies and models
• Large Language Models (LLMs) — Large language models
• Student Experience — Student experience with AI in higher ed
• Educational Development — Faculty development and AI readiness
• Academic Integrity — Academic integrity in an AI-capable student body
• AI Literacy — AI literacy for students and faculty
• Assessment Validity — Assessment validity in the age of GenAI
• Educational AI Policy — Educational policy on AI
• AI Regulation in Education — Regulation of AI in education
• Remote Proctoring — Remote proctoring and automated exam integrity
• Self-Directed Learning — Self-directed learning with AI
• Problem-Based Learning — Problem-based learning with AI
• Teaching — Evolving teacher role in AI classrooms
• Stakeholders — Umbrella: people and audiences in AI education (learners, teachers, designers, administrators, policymakers)
• Arts, Design and Media Education
• Student Support and Success — advising, outreach, referral, and support allocation — the institutional side of student success

Connected Articles

• How colleges and universities are adopting AI in higher education: a global comparative analysis — Chan (2026): AI Institutionalization Index across 100 top-ranked universities; discourse outruns structure, adoption independent of prestige
• Faculty Readiness for AI-Supported Teaching and Scalable Online Program Delivery in Higher Education: The EPIQ-AI Framework for Epistemic Integrity — EPIQ-AI Faculty Readiness Framework
• AI in the Wild: A Large Scale Analysis of Authentic Interactions of College Students with Generative AI — AI in the Wild: College Student AI Use
• A framework for institutional change in the age of AI — Institutional Change in the Age of AI
• Generative AI Availability, Grades, and Student Satisfaction at a Large University — GenAI Availability and Student Satisfaction
• 'A bit of chaos and madness': The AI Assessment Scale and the work of assessment reform — AI Assessment Scale and Reform
• The University AI Didn't Replace: Rethinking Universities in the AI Era — Rethinking Universities in the AI Era
• AI adoption among teachers: Insights on concerns, support, confidence, and attitudes — AI Adoption Among Teachers
• "AI should help them learn, not learn for them": University staff perspectives on the role of Generative AI in education — Staff perspectives on GenAI in higher education
• Exploring Organisational Drivers and Innovation Attributes of Artificial Intelligence Adoption in Higher Education — Organizational drivers of AI adoption in higher ed
• Is It Ethical for Teachers to Use AI for Student Feedback? — AI in student feedback: ethics
• Beyond Detection: Redesigning Authentic Assessment in an AI-Mediated World — Beyond detection: redesigning authentic assessment in an AI-mediated world
• "It is a temptation to get it to do the work…" Student experiences of navigating the generative AI landscape in UK higher education: A cross-institutional survey with international comparison — GenAI experiences among UK higher-education students (survey)
• AI skills for college graduates: Exploring how instructors and employers prioritize AI skills differently — Instructors vs. employers on AI skills for college graduates
• Faster Completion, Less Learning: Generative AI Reduced Study Time on Math Problems and the Knowledge They Build — 26.9% study-time decline among college students
• Artificial Intelligence in UK Higher Educational Policy and Institutional Decision Making — UK higher-education AI policy
• Is using artificial intelligence tools for academic work cheating? Student perceptions, ethics, and the impact on academic performance and critical thinking — AI tools, academic work, and cheating
• Assessing the Quality of AI-Generated Exams: A Large-Scale Field Study — Assessing the quality of AI-generated exams: a large-scale field study
• Reconsidering the use of oral exams and assessments: An old way to move into a new future — Reconsidering oral exams as authentic, AI-resistant assessment
• Assessment twins: An approach for strengthening assessment validity in the age of generative AI — Assessment twins for strengthening assessment validity in the age of GenAI
• Students' Perceptions of Artificial Intelligence Tools for Study Productivity and Learning: An Exploratory Survey Study — Students' Perceptions of Artificial Intelligence Tools for Study Productivity and Learning: An Exploratory Survey Study
• Understanding what good education is: a phenomenographic investigation of university teachers' understandings of AI — phenomenographic study of university teachers' conceptions of AI
• AI for education: The digital transformation of a liberal arts institution – implementation at Lingnan University — Digital transformation of a liberal arts university toward a research-intensive model in the GenAI era (Qin 2026)
• Mapping artificial intelligence integration in higher education: A systematic review using the FACETS and SAMR frameworks — Systematic review mapping AI integration in higher ed via FACETS + SAMR frameworks (AlSheikh et al. 2026)
• Navigating uncertainty: university teachers' experiences and perceptions of generative artificial intelligence in teaching and learning — University teachers' experiences and perceptions of GAI: vulnerability, rethinking assessment, student learning at risk (Farazouli et al. 2026)
• GPT-4 feedback increases student activation and learning outcomes in higher education
• Automated Grading of Open-Ended Questions in Higher Education Using GenAI Models
• Innovative assessment and grading practices in higher education: A critical exploration for management educators
• Shaping responsible GenAI use in research through AI literacy-oriented guidelines: Insights from postgraduate students — Postgraduate researchers' GenAI use in the research workflow and AI-literacy-oriented guidelines (Dai & Chan 2026)
• Engagement Intensity as a Learner-Modeling Signal for Adaptive AI Ethics Instruction — Engagement intensity as a learner-modeling signal for adaptive AI ethics instruction (Oh, Talton & Bui 2026)
• Smaller, Younger, and More Impactful: How AI-Assisted Writing Transforms Research Teams — AI-assisted writing shifts research teams toward smaller, junior-leaner, highly cited collaborations (Wang et al. 2026)
• StudentBench: AI and human tutoring yield equivalent GRE learning gains — StudentBench: AI and human tutoring yield equivalent GRE learning gains
• Literacy demands placed on students in student–GenAI entanglements in higher education — Assistant and Enabler forms dominate higher-education GenAI research; the marginal entanglement types stay rare for institutional rather than technical reasons (Sun, Dohn & Rehm 2026)
• Students at Stake: Risks of AI Deployment in Higher Education — Student-side map of AI deployment across admissions, financial aid, student services and education delivery, with a four-question procurement screen (Student Defense 2026)
• AIRRI: Towards a Human-Readable, Machine-Resistant Reading Interface — Perturbed-image assignments raised LLM completion failure from 7.4% to 55.6%: content design as an integrity route
• On AI glasses and wearable AI in assessment — AI glasses and hearing aids defeat an in-room exam ban: dual transparency with no gaze or posture cue
• Artificial Intelligence in Higher Education (2019–2026): A Systematic Review of Applications and an Evidence-Derived Multi-Axial Taxonomy — A functional-domain taxonomy of higher-education AI applications, 58.4% of which fall outside the 2019 categories
• AI and higher education: teachers' perceptions of students' ethical tensions in online pedagogical methods — Write method-specific guidance, not one campus-wide AI policy
• A Multidimensional Analysis of AI Literacy Determinants: External Resources, Digital Skills, and Psychological Profile Among University Students — Perceived resources and teacher support explained 29% of AI-literacy variance; demographics did not
• Exploring University Faculty's AI Well-Being: A Structural Equation Model of Social Supports, AI Literacy, and Technological Self-Efficacy — Fund AI literacy, not just tool confidence, for faculty