Decide your approach to generative AI
What approach to generative AI should you implement in your course: total prohibition, free usage without supervision, or somewhere in between? All have advantages and disadvantages, so it depends on the learning activity. The approach should be aligned with the learning outcomes and the type of knowledge or abilities taught and assessed.
We will explain three different approaches, spanning from total prohibition to free usage without supervision. Note that it is possible to use a combination of these approaches or use different approaches for different parts of your course. The most important point is to align the approach to the learning outcomes and to the type of knowledge or abilities that will be taught and assessed. For example, maybe your students should not be allowed to use generative AI without supervision during an exam if most learning outcomes are based on factual knowledge.
Approach 1: Prohibit all use of generative AI during examination
Prohibit all use by controlling the setting in which your students' knowledge is assessed. An advantage of this approach is that you can test factual knowledge. You can be certain the answers are the students’ own and what you assess is their knowledge and nothing else. However, this approach might be challenging and time-consuming to enforce. Also, it does not help students prepare for a possible future where AI is a working tool.
Use this approach for learning outcomes that are mostly factual and not specific to a course setting or content.
The easiest way to achieve this is to have a proctored examination, but there are several other ways. Read more about how to Hinder and detect prohibited use of generative AI .
Approach 2: Assume it will be used and adapt your course
Assume that your students will use generative AI to study, learn and possibly to produce submissions. Adapt your teaching and examinations accordingly. Teach your students how to use generative AI as a powerful tool in relation to your subject. Help them take advantage of its strengths while compensating for the weaknesses.
This approach might be time consuming if you must rewrite assignments, exercises or examinations. They must incorporate generative AI tools while also training students' abilities such as reflection, source criticism and analysis.
Use this approach for learning objectives where students should reflect, analyse or discuss a specific topic. Most Generative AIs are unable to produce useful material of this type independently. In addition, there are several potential benefits that students might get from learning to use generative AI in a purposeful way. Read more about how to Personalise learning with generative AI .
For now, we recommend the approach “allow – do not force”. Allow your student to use generative AI, but do not make it mandatory. This is due to GDPR and legislations discussions that are still in motion. We do not want to enforce the use until we are certain that it is legal.
Approach 3: Let students use it freely without supervision or support
Assume students will use generative AI to study, learn and possibly to produce submissions. Allow them to use it in whatever way they want to, to what extent they see fit. This approach will be choosen for you if you do not make an active choice.
The advantage with this approach is that you don’t have to change your teaching methods or your examination. However, for unsupervised exams there is a risk that more students than before will present submissions as their own that they haven’t produced themselves. This will cast doubts on the validity of the grades you give in the course. You will also miss an opportunity to teach your students how to use generative AI in an academically honest way.
This approach is not recommended. The reason we present it here is to inform you of the risks of not making an active choice regarding generative AI. For some courses this approach might work well enough, probably by coincidence. But we always recommend you analyse your course content, learning objectives and current assessments to see if they are aligned with each other. Thereafter you should actively decide your approach regarding generative AI in your course.
Further reading
Adapting assignments:
My class required AI. Here's what I've learned so far (oneusefulthing.org) .