Note: This page contains content and documents (including PDF) that have been automatically translated using Deepl.

23.4 What role does the use of AI play in the grading of exams (Section 12(4) MRVO)?

According to Section 12(4) of the MRVO, “exams and types of exams enable a meaningful assessment of the learning outcomes achieved” and are designed to be “module-based and competency-oriented.” The MRVO thus gives higher education institutions flexibility in the design and further development of their examination formats.

It is therefore up to the higher education institutions to design their examination systems in such a way that, even under the conditions under which AI is used, a meaningful assessment of the learning outcomes achieved remains possible and that exams fulfill their purpose in accordance with Section 12(4) of the MRVO. Depending on the qualification goals, , both the exclusion and the permitted use of AI may be appropriate. ​

The accreditation process should examine whether the assessment system—even when AI is taken into account—is generally suitable for providing a meaningful evaluation of learning outcomes. In doing so, it must be clearly defined which forms of AI use are permitted and what requirements are placed on students.

To ensure that proof of competence is valid in accordance with Section 12(4) of the MRVO, different measures may be appropriate depending on the study programme. These include, for example:

  • clear guidelines on the use of AI (e.g., permissibility, documentation requirements, declarations of autonomy),
  • Exam formats that require the application of knowledge,
  • the combination of different types of exams, such as written and oral components.

This list is illustrative and not exhaustive. If this results in multiple examinations that may need to be passed separately, this is permissible under accreditation law, provided the requirements of § 12(5)(1) MRVO are met (see FAQ 21, “examination concept”).