10 Guidelines for Assessment Practice in a GenAI Environment

This post was updated for links and terminology in October 2023.

Generative artificial intelligence (GenAI) tools, such as but not limited to ChatGPT, are increasingly available with quickly advancing capabilities.  Proper and ethical use is important for transparent and valid assessment.  The following are 10 general guidelines for educators.

To prepare

1. Learn about relevant, new GenAI technologies and their applications in contexts of interest.
2. Discuss approaches and emerging practices with disciplinary colleagues.
3. Design assessment to meet course learning outcomes in ways that include acceptable uses of GenAI tools and/or reduces the likelihood of unacceptable uses.

During the course

4. Explain how students should and should not complete their assessments and the reasons for these expectations. Be explicit about the tools, services, and actions of concern.
5. Emphasize the learning value to students of completing the assessment as per expectations.
6. Equip students with the information and experiences they need to develop the skills and knowledge required to meet the expectations. Normalize help-seeking.
7. Reassure students that their questions about the assessment expectations are welcome. Academic misconduct is often a mistaken response to uncertainty and time pressures.

Then, if you suspect or identify academic misconduct

8. Follow up proportionately and consistently to suspected academic misconduct. Unpermitted assistance and/or misrepresentation of authorship may be specific concerns.  No GenAI detection tools are approved for use at USask at this time.
9. Protect Copyright. Do not submit the work of others (including work of students) to third-party services, such as GenAI detection tools, without permission.
10. Reflect on results of the approaches used and plan for next time.