From the Student Success edition of November 8, 2026
Ctech (5 Nov): Hebrew University committee says take-home work is no longer a valid assessment, recommends oral exams and in-class presentations while warning of fairness costs
Hebrew University of Jerusalem · Higher education assessment design · Israel
- Publisher
- Ctech by Calcalist
- Original publication
- Ctech (Calcalist), 5 November 2026
- Source retrieved
- 2026-11-10
What happened
Ctech reports that a Hebrew University committee, appointed by the rector and chaired by the vice-rector for academic affairs, concluded that unsupervised assessments such as take-home exams and homework are susceptible to AI use, which renders the assessment invalid. Its proposals, adopted by the university Standing Committee, include requiring in-class presentations of seminar papers followed by questions, more oral examinations, fewer unsupervised take-home assignments, more in-class tests and exercises, and participation grades. The committee also distinguishes AI used to improve thinking from fully AI-written submissions, which it treats as plagiarism, notes that supervised exams limit assessment of important skills, and warns that oral exams risk unfairness for non-native speakers, students with speech difficulties and shy students. Each academic unit must review its assessment methods and instructors need training. A January student-union survey found 87% of students use AI frequently.
Why it matters
Offers an international, institution-level decision to move away from unsupervised assessment, with explicit fairness and standardization costs that U.S. campuses weighing oral or in-class formats must also handle; transfer is limited by differences in context and scale.
Evidence and measured results
Secondary news reporting of a committee report and one student-union survey; no pilot results, validity data or implementation outcomes. The report itself was not reviewed.
Limitations and uncertainty
Non-U.S. institution; news report only; survey is a student-union survey with unspecified sampling; effects on learning or fairness not measured.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-11-10; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Unsupervised assessments are judged unreliable under AI, but supervised alternatives cost academic scope and risk unfairness.
- Stakeholders
- Rector or provost
- Deans and department chairs
- Teaching and learning center
- Student union
- Accessibility office
- Discovery questions
- Which courses rely mainly on unsupervised take-home assessment?
- How will oral exams be standardized across assessors?
- What accommodations are planned for language and speech differences?
- Value hypothesis
- Interpretation: redesigned, partly supervised assessment may restore evidence of student ability if standardization and accommodations are handled; not shown here.
- Potential engagement
- Assessment-redesign program: course inventory, pilot of oral or presentation formats, rubrics and lecturer training.
- Unsupported claims
- Improved assessment validity
- Reduced AI misuse
- Better learning outcomes
Pre-sales engineering
Role takeaway
- Fit
- Assessment operations tooling for oral and in-class formats.
- Architecture and integration
- Scheduling and room booking, rubric tools linked to gradebook, optional recording storage.
- Prerequisites
- Course inventory by assessment type
- Rubric templates
- Accommodation procedures
- Constraints
- Oral exam standardization is difficult
- Accessibility disadvantages for some students
- Non-U.S. context
- Security
- If oral sessions are recorded, set consent, access and retention rules.
- Proof of value
- Pilot oral format in several courses and compare assessor consistency
- Test accommodation workflow
- Measure staff time per student
Delivery
Role takeaway
- Work
- Review assessments by unit, pilot formats, train lecturers and publish rubrics.
- Dependencies
- Department chairs
- Teaching center
- Accessibility office
- Scheduling office
- Ownership
- Vice-rector for academic affairs with unit heads
- Skills and adoption
- Lecturers need skills and examples to design presentations, oral exams and in-class tasks.
- Governance checkpoints
- Fairness review for language and speech differences
- Unit curriculum review sign-off
- Annual assessment-method audit
- Acceptance criteria
- Proposed: all units complete assessment review by a set date
- Proposed: assessor agreement measured on pilot orals
- Proposed: accommodation requests resolved within a target time
- Risks
- Staff workload and scheduling strain
- Unfairness to non-native speakers
- Loss of skills assessment from returning to exams
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Shifts assessment from LMS submission toward in-person and oral formats, with implications for scheduling, rooms, recording and rubric tooling rather than AI software.
Governance
Who approves, reviews and stays accountable for outcomes?
Unit-level mandatory review of assessment methods, with a distinction between AI-assisted thinking and AI-written submissions, creates a policy basis for plagiarism handling.
Security and privacy
What data, permissions and controls need testing?
Oral assessments may involve recording or storing student speech; the article does not address privacy or retention.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Committee flags disadvantage for non-native speakers, students with speech difficulties and shy students; instructors need new assessment skills and likely more time per student.
Procurement
What should contracts, pricing and exit terms secure?
Little direct procurement impact; may increase demand for assessment-design training, scheduling and rubric tools.
Operating model
Which teams own the service once it runs?
Rector-level mandate pushes decisions to academic units, which must review curricula and train lecturers, shifting workload to departments.
Publication history
- 2026-11-08Student Success · Issue 633 resources
Stable resource ID: ctech-hebrew-university-committee-oral-exams-ai-assessment-20261105