Abstract:

Active, formative assessment at scale is one of the most persistent challenges in higher education courses. This presentation describes an ongoing development project spanning over a year, in which a university lecturer in International Business designed, tested, and iterated on a set of AI-powered teaching tools — working within the constraints of an existing LMS environment, without the need to involve institutional IT support, and within the boundaries of GDPR.

Three components are described: First, a prompt-based question generation pipeline that produces large, ready-to-upload question libraries for any subject — from business ethics to biology — in a fraction of the time manual question writing requires. A single prompt generates multiple test versions simultaneously, each with adjustable difficulty, varied question formats beyond multiple choice, and enough variation between versions to make answer-sharing between students structurally ineffective. The resulting files upload directly to Brightspace and compatible LMS platforms without technical intervention. What previously required hours of individual question writing becomes a repeatable, scalable process any teacher can run independently.

Second, a set of AI powered games from competitive team-based in class engagement and practice, to individual self study and formative assessment, using case studies and scenarios modelled on real company data from Dutch and international organisations, piloted with students and iterated based on observed engagement.

Third, an AI powered individual interviewer and tutor tool that assesses open-ended student reasoning against a rubric and provides immediate written feedback — currently in testing ahead of broader use.

Initial pilot observations from fellow lecturers suggest that using AI in quiz generation helped improve efficiency and productivity, but also helped reduce manual administrative tasks, with reduction of mistakes, and inconsistencies. Students who tried a few different games during an AI elective course found them engaging and helpful, easy to use and noted that AI-generated feedback provides sufficient specificity to function as a genuine learning tool rather than a novelty.
The presentation includes a live demonstration and invites audience participation in a short case activity.

Keywords:

AI, assessment, engagement, technology, activities, higher education, games.