While your AI policy should specify whether students are allowed to use AI in the course assessments and to what extent, since AI use is difficult to detect, students may still make unauthorised use of it. Such unauthorised use may undermine the validity of assessment if it makes it impossible to establish whether students achieved intended learning outcomes. Moreover, some methods are more vulnerable to AI cheating than others, as represented in the image below.

Given the fast development of AI, “out-designing” AI from take-home assessments is becoming increasingly challenging. The table below summarises three possible approaches concerning AI and your assessments, the modalities each choice entails, and the contexts in which they may be most appropriate. Go to the related hub pages for a more in-depth discussion of these approaches, along with practical ideas and tips for their implementation.

Choice	Modality	Best for…
Choosing an AI-proof assessment method	On-campus assessment	Assessing the acquisition of essential knowledge and skills.
Implementing measures to limit the impact of AI on the validity of take-home assessments	Take-home assessment	Ensuring that you can test students’ original contribution even if AI is used.
Incorporating AI into assessment tasks	Take-home and on-campus assessment	Make AI part of the learning and assessment process.