Universities can secure the parts of assessment that count towards final marks without scrapping coursework, argues higher-education specialist Chris Rust. His practical alternative is to keep coursework for learning, then test students’ understanding under controlled conditions with tasks that reward applying knowledge, not simply recalling it.
Watch Desk analysis
What happened
Rust, emeritus professor of higher education at Oxford Brookes University, is responding to coverage of the Office for Students’ advice on AI and assessment. He argues that the regulator’s call for secure assessment should not be read as a return to traditional exams or a reason to abandon formative coursework.
His suggestions include ungraded work for self-assessment and peer discussion, followed by exams that test how students use what they have learnt. Students might analyse an unseen article, advise a client on a case, or build on coursework from a different perspective. He also makes the case for more vivas, where students explain and defend their work.
Rust sets out this argument in The Times Higher Education. These are his proposals, not evidence that any one assessment format will work equally well across subjects.
Why it matters
AI makes it harder to know whether submitted work reflects a student’s own understanding. But removing coursework could also remove opportunities for feedback and practice. Rust’s distinction offers a way to tackle both problems: use coursework to support learning, and make the final assessment show what the student can do.
That means the choice need not be between trusting every take-home essay and returning to timed recall tests. Open-book exams and follow-up questions can test reasoning and application, while still requiring students to demonstrate their own knowledge.
Our read
Rust is right to challenge the idea that “secure” has to mean “old-fashioned”. The strongest part of his argument is the focus on what an assessment is meant to measure. A viva or a well-designed applied exam can make borrowed or AI-generated work harder to pass off as understanding, without treating coursework as educational dead weight.
The catch is practical: fewer, better assessments and individual oral examinations still take staff time to design and deliver. Universities should try approaches that fit their subjects, and explain how students will be assessed and where AI use is allowed.
What to watch
- Whether universities separate formative coursework from the work that determines final marks.
- Whether applied, open-book exams and vivas appear in new assessment plans.
- How institutions provide enough staff time to run those assessments well.
Discussion spark: Should universities keep coursework for practice and feedback while reserving final marks for controlled assessments, or can coursework itself still be a fair measure of learning when students have access to AI?
Sources and evidence
- Source update (18 November 2026, 00:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.