News
Cadmus convenes a working group to develop a new approach to oral assurance for secure assessment

Share with colleagues
Download the full Case Study
Take an in-depth look at how educators have used Cadmus to better deliver assessments.
Cadmus is excited to announce the formation of a Working Group, bringing together universities and academic leaders from APAC and the UK to collaboratively design a new approach to oral assessment for an AI-enabled world.
As generative AI reshapes how students produce assessment work, universities are increasingly seeking assessment approaches that can confidently evidence genuine understanding — without defaulting to surveillance, detection, or blanket oral exams. Across both regions, partners highlight a shared challenge: oral assessment is widely trusted, but is difficult to operationalise at scale in real university conditions, and needs to be carefully considered to support students in the best way possible while accommodating for diverse cohorts.
Through early discussions, institutions pointed to a consistent set of tensions — balancing learning and assurance, scaling across large cohorts, managing workload, supporting equity and accessibility, and meeting evolving regulatory and policy expectations. These conversations reinforced the need for a learning-centred, proportionate approach to oral assessment, embedded within the assessment journey rather than bolted on as a last-resort integrity check.
Cadmus Oral Assessment is being developed to address this gap.
The solution will be a learning-centred, educator-controlled oral assurance layer, embedded directly into existing assessment journeys. Rather than treating oral assessment as a policing mechanism, Cadmus is looking to introduce targeted, proportionate oral moments where explanation, reasoning, and knowledge ownership matter most — strengthening confidence in learning outcomes while remaining scalable, inclusive, and defensible. Most importantly, using Oral assessment as just one layer in a broader assessment ecosystem.
The solution is being shaped around the realities universities shared, including:
- Using oral assessment as a selective assurance layer, not a universal requirement — for example as an oral defence of written assessments or exams, as part of group assesments (including as a prepatory exercise), an integrity trigger, or an observed assessment within a broader programme design.
- Supporting asynchronous and digital oral assessment at scale, particularly for large undergraduate cohorts where traditional live oral assessments are challenging.
- Embedding oral assessment within learning journeys, linking to previous assessments, with low-stakes, formative experiences early on that build confidence, and more dialogic, human-led oral assessment later where judgement and synthesis matter.
- Designing oral assessment that are fair, transparent, and accessible, with clear expectations, rubric-aligned marking, and support for reasonable adjustments.
- Using AI carefully and transparently to support preparation, prompting, and scalability — while keeping academics visibly in control and firmly in the loop.
The Working Group will play a central role in shaping this solution, ensuring oral assessment reflects real academic workflows, diverse disciplinary needs, and practical implementation realities.
What this means for our existing partners
For our global partners, this innovative solution marks the beginning of a new era in higher education.
- Cadmus Webinars: We look forward to collaborating with academic leaders through our series of webinars, bringing together educators from across disciplines and regions to shape the future of secure, learning-centred oral assessment. Webinar registration coming soon.
By working in close partnership with universities from the outset, Cadmus aims to deliver a oral assessment solution that institutions can confidently adopt.
Thank you to our partners from:
- Australian Catholic University
- Charles Sturt University
- Deakin University
- Edith Cowan University
- Flinders University
- La Trobe University
- London School of Economics and Political Science
- Massey University
- Queen Mary University of London
- RMIT University
- The University of Auckland
- The University of Queensland
- The University of Waikato
- The University of Glasgow
- The University of Melbourne
- University College London
- University of Newcastle
- University of New South Wales
- University of Southern Queensland
- University of St Andrews
- University of Tasmania
- University of Warwick
- University of Western Australia
- University of Notre Dame Australia
- Victoria University
Category
Company
Product update
More News
Load more
AI
Leadership
Why “AI-proofing” assessments misses the point—and what to design for instead
This article challenges the idea of “AI-proofing” assessment, arguing that restriction and detection strategies fail to address the real issue: assessment designs that no longer produce reliable evidence of student capability. Nick Bareham explores how universities can instead design assessments that make thinking, judgment, communication, and applied understanding visible—the very capabilities employers are increasingly demanding in an AI-enabled world.
Nick Bareham, Chief Growth Officer, Cadmus
2026-06-09

AI
Leadership
Why AI demands a redesign of assessment—not just a policy response
This article argues that AI has exposed a deeper structural weakness in higher education assessment: the reliance on outputs as proxies for genuine learning and capability. Brigitte Elliott explores why policy and detection tools alone cannot solve this challenge, and why institutions must redesign assessment to make student thinking, reasoning, and understanding visible.
Brigitte Elliott, Co-CEO, Cadmus
2026-05-26

AI
Leadership
The degree has always been a signal. AI is changing how we verify it.
This piece explores how AI has fundamentally disrupted the university degree as a trusted signal of capability, decoupling assessment outputs from genuine student understanding. Herk Kailis argues that institutions must move beyond AI detection tools, toward approaches that require students to demonstrate authentic understanding in real time.
Herk Kailis, Founder & Co-CEO, Cadmus
2026-05-11