Information for students

When a student requests an explanation, the examiner can use an AI tool that suggests a first draft of the explanation based on:

• the exam paper
• the examiner's guidelines
• the student's answer
• the student's grade
• the examiner's own notes

The examiner then reads through the draft, edits and quality assures the text, and sends the final explanation to the student.

The examiner is always academically and legally responsible for the content of the explanation.

If the AI tool has been used to create the draft, the examiner must inform the student of this.

Every year, UiA receives around 10,000–12,000 requests for explanation, and many lecturers find that the work takes a lot of time without the learning outcomes for students always being proportional to the effort.

The aim of the project is to:

• ensure faster and more comprehensible explanations,
• provide more consistent quality in feedback, and
• free up time for more academic guidance and learning activities.

The project is being carried out in collaboration with the Student Organisation in Agder (STA). The students' experiences and perspectives are central to the development work to ensure that the solution is perceived as fair, useful and trustworthy.

As part of this collaboration, a survey has also been conducted among UiA students in autumn 2025.

The survey maps students' expectations and experiences with explanations and provides insight into how these can be improved – both with and without the use of AI.

The findings of the survey are included in the project’s final report.

STA leader Magnus Ellefsen Moen says the following about the project:

This is an incredibly exciting project. Many students have probably felt that explanations can be delayed and that what is written in them does not quite match what they hoped for. I truly believe that this project will improve this situation. We have previously been sceptical about using KI for this particular purpose, but the student perspectives and input we provide are included and well received. This is greatly appreciated by myself, STA and the students. We are very pleased to be able to contribute to the project and look forward to continuing to work on this topic.

All use of data takes place within a secure framework. Assignments, examiner guidelines and answers are processed in a closed and secure environment, in accordance with applicable privacy regulations (GDPR) and the university's internal guidelines.

No one other than those who already have access to this data in a normal examination process – such as examiners, subject teachers and the examination office – will have access in connection with the use of the tool.

No data is used to train the AI model, and the information is not shared outside UiA.

This means that the tool does not store or reuse information from exam answers, and that students' work does not leave the university's secure systems at any time.