AI and examination (120 min)
Increased access to powerful AI brings new opportunities and challenges for how we approach assessment and evaluation. In this workshop, we will review research on AI and assessment and highlight strategies to minimize the risk of AI‑related misconduct in examinations. You will also get the opportunity to discuss and reflect together with colleagues at KTH on generative AI and its impact on examination processes.
Purpose
You, as a teacher, will get the opportunity to discuss and reflect with other colleagues at KTH about generative AI and its impact on examination processes. we will review research on AI and assessment and highlight strategies to minimize the risk of AI‑related misconduct in examinations.
Preparations
To effectively participate in the workshop, it is important that you have done the following before the session.
-
You, as a participant, must have access to a text‑based generative AI tool. All employees at KTH have access to Microsoft Copilot Chat (the recommended AI tool at KTH). Information on how to access the service can be found on the webpage: AI services: Microsoft Copilot Chat .
-
You should have briefly read through OpenAIs guide för “Prompt engineering” (eng, platform.openai.com) which contains strategies for getting better results from large language models.
-
You should have tried using generative AI, for example Copilot, and the strategies you have read about above to solve one or more examining assignments from your own courses.
Time required
120 min.
Learning objectives
After completing the workshop, participants should be able to:
-
Identify opportunities and challenges related to generative AI and assessment, for example, in the design, execution, or grading of examinations.
-
Reflect on how you, as a teacher, can design examinations that guide students away from AI‑related misconduct.