AI & Assessment
Safeguarding the value of our degrees
If students can pass assignments or even entire courses through the unauthorized use of genAI, we have no way to ensure that they have actually mastered the learning outcomes and, ultimately, the exit qualifications for their degree. Although most students do not necessarily want to use shortcuts to get their degree, we also know that many of them already use genAI for both formative and summative assessments. There is no way of preventing this with unsupervised assessments, and we cannot rely on detection tools to determine whether a text was written by genAI. This has a considerable impact on the way we organize our teaching and assessment. We should keep in mind that the goal is not to police students but to create an expectation of trust, to stimulate the responsible use of AI, and to help students engage with the learning process appropriately.
Long-term impact and current limitations
In the long-term, the rise of genAI will require us to evaluate and revise our current teaching and assessment thoroughly. As AI affects both teaching practices and society at a larger scale, this will involve a wide range of changes, from adapting learning outcomes to changing how those outcomes are assessed. Ultimately this will involve actively teaching students how to use genAI responsibly and incorporating it into our courses. However, we cannot begin to do this until genAI is compliant with privacy regulations (GDPR) and safely accessible for both staff and students. Only then can its use be approved by the university.
Short-term adjustments
For now, we can focus on the short-term impact of genAI on assessment. You will find further information on the two pages below: genAI and assessment at course level, and genAI and written assignments.
