Oral Defenses (Vivas) as the New Standard for Academic Integrity

What You Need to Know

Oral defenses (viva voce) are being adopted by universities worldwide as the most effective way to verify authentic student work in the AI era
A 5–10 minute oral conversation after a written submission changes the entire dynamic of assignment integrity
Stanford’s Academic Integrity Working Group recommends oral exams and in-class formats for high-stakes assessments
Research in Frontiers in Education (2024) found viva voce exams among the most effective AI-resistant formats
UC San Diego’s NSF-funded study implemented oral exams across 32 classes, finding they also improve learning outcomes
Many UK and Australian universities now include vivas as a mandatory component of their academic integrity policies

Why Written Submissions Are Losing Their Grip on Academic Integrity

For over a century, the essay and the written exam have been the default forms of assessment in higher education. They test research skills, critical thinking, and written communication — all valuable intellectual capacities. But now we face a problem that no assessment method predicted: generative AI can produce coherent, researched-looking essays in seconds.

It’s not about the essay being a “bad” assessment format. Essays remain excellent for building analytical writing skills. The problem is simpler: a written document can be outsourced, ghost-written, or AI-generated without the instructor knowing. Even the best AI detectors produce false positives — one researcher tested their own 2006 dissertation through a popular AI detector and it flagged 39% as AI-generated. Courts in several countries have ruled against institutions that penalised students based on AI detection scores alone.

When written work alone can’t reliably verify that a student produced their own thinking, we need a method that must test the person, not just the product.

That’s exactly what an oral defence — also called a viva voce examination — does.

What Is an Oral Defense (Viva Voce)?

An oral defense is a structured conversation between a student and one or more examiners, in which the student explains, justifies, and responds to questions about their own work. It is not a new concept. It has been the cornerstone of PhD examinations for centuries and is standard in professional fields like medicine. What’s happening now is different: oral defenses are moving down from postgraduate to undergraduate and even taught-postgraduate level as a routine academic integrity safeguard.

At its simplest, an oral defense looks like this:

The student submits written work (essay, project, report, or portfolio)
The student then has a 5–30 minute live conversation with an examiner
The examiner asks follow-up questions about the student’s methodology, reasoning, and conclusions
The student must explain their own thinking on the spot

No AI, no search engine, no chatbot can sit in that room. The student must demonstrate real-time comprehension of their own work.

How Oral Defenses Deter Academic Misconduct

Here’s the key insight that’s driving the current shift: oral defenses don’t need to catch cheaters to work. They just need to make cheating the path of least resistance.

When a student knows there’s a viva component, the incentive structure changes. A student who used AI to write a paper still has to:

Explain their choices under live questioning
Demonstrate they actually performed the research
Respond to follow-up questions about details they couldn’t have fabricated

The path of least resistance shifts back towards doing the work well enough to talk about it. This is what researchers call an “AI deterrent” rather than “AI-proofing” — it diminishes the appeal of complete cognitive offloading without claiming to make cheating impossible.

Newcastle University’s 2026 viva implementation with 60 undergraduate students found that “students judged the viva surprisingly user-friendly, and almost all valued the chance to talk at length about projects they had chosen and cared about.”

UC San Diego’s study (funded by an NSF grant) found that oral exams reduced the perceived need to cheat simply because students believed the course was more honest. The “having to cheat to keep up” motivation — one of the biggest drivers of academic misconduct — was directly weakened by the knowledge that an oral component existed.

How Oral Defenses Improve Learning Outcomes (Beyond Integrity)

Here’s what many instructors don’t expect: oral defenses often improve learning outcomes beyond just catching dishonesty. Multiple studies across engineering, business, and the humanities have identified consistent benefits.

They reveal real understanding vs. memorisation. At UC San Diego, oral exams exposed students whose learning strategies were focused on “mere memorisation of processes and recipes.” Students who hadn’t truly understood their material often couldn’t explain their written answers when questioned. But crucially, the oral format also revealed students who had performed well on written exams but struggled with conceptual understanding — a blind spot written assessments never caught.

They lift attainment for weaker writers. Oral formats tend to produce higher overall attainment, particularly for students whose natural strengths lie outside writing. A student who can articulate ideas clearly but struggles with academic prose often raises their mark through the viva component.

They create genuine engagement with feedback. At Newcastle University, students received their written feedback before the viva and had two weeks to prepare. Comments that might otherwise have been “skimmed and filed” were instead read closely and used actively in preparation. Feedback became “something to use rather than merely receive.”

They strengthen academic integrity culture. When students believe their course values authentic understanding over polished writing, they internalise that standard. The perception alone is powerful — students who trusted that the assessment was honest were less likely to feel pressured into dishonesty.

Why 2026 Is the Turning Point

Three signals converge this year:

1. Institutional policy adoption. UK and Australian universities are embedding vivas into their formal academic integrity policies. Middlesex University’s Academic Integrity and Misconduct Policy (2025-2026) mandates vivas when there are “suspicions of poor referencing, contract cheating, or AI misuse.” Several institutions now treat oral components as standard capstone assessment rather than a remedial response.

2. Research evidence base. Multiple peer-reviewed studies — including Frontiers in Education (2024), Studies in Higher Education (2020), and Assessment & Evaluation in Higher Education (2012 and 2026) — now confirm viva voce effectiveness across disciplines. Stanford’s Academic Integrity Working Group (2024-2025) explicitly recommends oral exams and in-class formats.

3. The detection arms race has lost. Every major university I’ve spoken to agrees: AI detection tools are unreliable, biased, and legally risky. The conversation has shifted from “how do we detect cheating?” to “how do we design assessments that don’t require detection?”

Practical Implementation: What It Actually Looks Like

Here’s how you structure an oral defense without overhauling your entire module:

Format

Duration: 5–15 minutes per student (shorter is sufficient for verification; longer allows deeper exploration)
Number of examiners: 1–2 (a single instructor is enough for basic verification; two examiners reduce bias and spread workload)
Mode: In-person or video call (online oral exams are viable and increasingly common)
Student preparation: Allow 1 side of A4 notes; let students prepare 3 questions about their own work

Question Design

Effective oral defenses don’t ask students to repeat their essay. They probe for ownership:

“What was the hardest part of this research, and how did you solve it?”
“If you had to change one methodological choice in your study, what would it be and why?”
“How does your conclusion relate to [specific source] that you cited?”
“What would you do differently if you started this project today?”

Weighting

You don’t need to make the viva the majority grade. Newcastle University found that when students voted on viva weighting (range 15–50%), the majority chose 30%. This was “high enough to make the review worth taking seriously, low enough to keep anxiety in check.”

Time and Workload

Yes, oral assessments are labor-intensive. That’s the tradeoff. UC San Diego found a ratio of one instructor or TA to 30 students as the maximum for implementation. At Newcastle, 60 students required 80 hours of staff time in total. The investment is real, but it’s often less than the time spent managing AI-related misconduct hearings and investigation.

What to Avoid

❌ Don’t turn the viva into an interrogation. Frame it as a review or dialogue — Newcastle University deliberately used “professional and personal review” rather than “examination” as the framing. This matters. Students approach a viva differently when they know it’s a chance to demonstrate their work, not a trap.

❌ Don’t ask students to repeat their written submission. If the examiner can ask the same questions the essay already answered, the oral component is redundant. Follow-up questions must probe beyond what’s on the page.

❌ Don’t skip scaffolding. Students who’ve never been assessed orally often experience anxiety. An optional preparation workshop (even with thin attendance) helps enormously. Letting students review written feedback before the viva creates a concrete reason to engage with assessment comments.

❌ Don’t use vivas as a first resort for suspected misconduct. They work best as a routine component for all students. When vivas are deployed only as a “catch cheaters” measure, students feel mistrusted. When they’re routine, the integrity benefit compounds.

When Oral Defenses Work Best

Not every course needs a full viva. Here are scenarios where oral verification adds the most value:

Scenario	Recommended Approach
Capstone/thesis projects	Full oral defense (15–30 min)
Group project accountability	Individual 8–10 min Q&A
Mid-semester high-stakes assignment	Short oral check (5 min)
Courses with prior AI concerns	5–10 min follow-up on every written submission
Large classes (100+ students)	Staged submissions + spot oral checks

A Note on K-12 and School Settings

Oral defenses are not limited to higher education. UC San Diego’s research team implemented them across multiple undergraduate engineering courses, and the principles apply at any level where authentic understanding must be verified. In K-12 settings, shorter oral checks (3–5 minutes) after project presentations or written reports can serve the same purpose: confirming that the student actually knows what they produced.

Where This Leaves Us

The shift toward oral defenses isn’t about returning to an older, less fair system. It’s about recognizing that writing has always been the wrong question to ask when you want to know if a student thinks.

A written essay that asks “Discuss the causes of the 2008 financial crisis” was already answerable without much original thought — AI has just made that answer even easier to produce. Oral defenses force students to do what no machine can replicate: think on their feet, defend their own choices, and explain their own reasoning to a human being.

The educators who will navigate this period best are not the ones who build higher surveillance walls. They are the ones who design assessments worth doing — tasks that require a student to be present, to think, to explain.

That’s not a concession to AI. That’s what good teaching has always looked like.

Related Guides

Frequently Asked Questions

Is an oral defense the same as an oral exam?

An oral defense typically focuses on a specific student project or written submission — the examiner questions the student about their own work. An oral exam tests a student’s knowledge of course material in general. Both serve as AI-resistant assessment formats, but oral defenses are uniquely powerful for verifying authorship because they probe ownership of the specific work submitted.

Can oral defenses be done online?

Yes. Online oral exams are increasingly common, especially since the pandemic accelerated remote assessment formats. Tools like EduLegit’s screen-recording and monitoring capabilities can support live online viva sessions while maintaining academic integrity safeguards.

What if I’m teaching a large lecture class?

The classic concern is workload, and it’s valid. But UC San Diego successfully implemented oral exams across courses with 150+ students using a team of instructors and TAs at a 1:30 ratio. Staged submissions — requiring topic proposals, annotated bibliographies, and rough drafts — can dramatically reduce the need for follow-up questioning during the viva, because students who didn’t do the work are often caught at the staging checkpoints.

How do I grade an oral defense fairly?

Use a rubric that assesses clarity, depth, and authenticity — not content accuracy alone. Grade the student’s ability to explain their reasoning, not whether their answer matches a textbook. Many educators use a simple three-point scale: demonstrates understanding (passes), partial understanding (borderline), cannot explain (fails). Keep the grading low-stakes to reduce anxiety.

Looking for tools that support both monitoring and assessment integrity? Explore EduLegit’s classroom management features to see how screen-recording, typing-activity monitoring, and real-time alerts can complement oral verification methods.