Guidelines for AI and Assessment: How do top-ranking universities respond to the challenges?

The latest developments in generative artificial intelligence are challenging the assessment practices of higher education institutions worldwide. In their article, Benjamin Luke Moorhouse and colleagues explore how higher education institutions respond to these challenges, looking at the guidelines of the world’s 50 top-ranking universities. Which areas do they focus on, and how do they support teachers?

Of 50 institutions, just 23 of them have published such guidelines that are publicly available on their websites. The guidelines that are available focus on three main areas: First, academic integrity, for example, the expectation that all the members „act with honesty, trust, fairness and responsibility“ (Tertiary Education Quality and Standards Agency, TEQSA) – in contrast with academic dishonesty and the policies to handle plagiarism. Second, advice on assessment design: Since generative AI can fulfill many traditional assessment tasks like written essays, there is a need for guidelines or suggestions on how to modify their assignment tasks – for example, by fostering creativity and critical thinking, incorporating contextual components or tasks that refer to real-world situations. Lastly, the authors identify communication with the students as the third important aspect, for example, communicating clear expectations about the use of AI and partnering with students about ethics and limitations of generative AI or the importance of originality, college learning, intellectual struggle, and process.

Looking at their results, the authors emphasize the importance of guidelines to support instructors to prevent a defensive approach to generative AI. Since the institutions have recognized that AI is here to stay, there is a need to adapt the assessments to it. Among other things, one important recommendation is to test the teachers’ assessments to see how the generative AI can accomplish the respective task. Another important aspect, following the author’s conclusion, is not to ban but to integrate GAI into assessment tasks and assess how students use it, since this is an important competency for future jobs. 

References
TUM’s guidelines and further information

Share it