Digital exam management: from scheduling to certificate

Examinations at German universities are regulated by law. The state's higher education act sets the frame; the programme's examination regulations set the details: assessment types, registration and withdrawal, resits, grading standards, deadlines for publishing results, reasonable adjustments and appeals. Digital examination management must reflect these rules — not the other way round.

Roles are clearly assigned. Lecturers create and grade their examinations. The examinations office manages registrations, dates and results, watches deadlines and issues certificates. Students see their own registrations, attempts and results. A good system technically prevents anyone from changing grades outside their role and logs every change.

Not every assessment type suits digital processes equally well. The overview shows typical types and what to watch digitally.

AI can lighten the grading load: it can pre-check free-text answers against a rubric, verify quotations and sources or flag anomalies. The decision must stay with a human. Article 22 GDPR gives data subjects the right not to be subject to a decision based solely on automated processing that produces legal effects concerning them — and an examination grade is such a decision.

The AI Act adds to this: systems intended to evaluate learning outcomes are, under Annex III point 3(b), in principle high-risk AI systems, with obligations for providers and deployers from the date set by Article 113. Universities should therefore design AI support so that suggestion and grade are separate fields and the lecturer's reasoning is documented.

Online exams raise the question of how to ensure that the right person takes the exam without unauthorised help. Camera proctoring is possible but demanding under data protection law. Transparency before the exam, data minimisation — such as single photos instead of continuous video — short retention and access only when a case arises are minimum standards. An alternative without recording should be considered.

Here too the AI Act applies: AI systems for monitoring and detecting prohibited behaviour of students during tests are classified as high-risk under Annex III point 3(d). Emotion recognition in educational institutions is in principle prohibited under Article 5(1)(f).

Graded examinations flow into the transcript of records: grade, credits, examiner, date. Students see their ECTS account and weighted grade average there. The transcript produces grade reports, the Diploma Supplement and finally the degree certificate.

For legal certainty: grades must be reasoned and traceable, students usually have a right to inspect their examination file, and there is an appeals procedure against grades. Digital graduation documents should be verifiable so that employers can check authenticity without asking.

Equal opportunity is a core principle of examination law. Students with a disability or chronic illness are usually entitled under the examination regulations to reasonable adjustments, such as extra time, a different form of assessment or technical aids. Digital examination management must be able to reflect this — individually, confidentially and without the reason being visible to others.

Digital examinations themselves must be accessible: operable by keyboard, readable with a screen reader, with sufficient contrast and without unnecessary time pressure caused by the technology. State universities are obliged under state law to ensure digital accessibility; for all universities it is a precondition for fair examinations.