Lesson 4 – Supporting Gen AI Use in Assessment

Giving students the opportunity to think critically and openly about their AI usage lays bare some uncomfortable truths for both students and teachers. It can lead both parties to question their assumptions and be surprised by what they find. Faculty members may discover that students actually learned something using AI; conversely, students might realize that their use of these tools meant they didn’t learn much of anything at all.

Marc Watkins, Chronicle of Education

As you design assessments with student use of Gen AI in mind, consider the following:

• Refine AI outputs. Ask students to improve AI-generated content and explain their changes. This builds editing skills and encourages reflection on clarity, tone, and accuracy.
• Document decision-making. Let students experiment with AI, but have them share why they used it, what prompts they tried, and how they evaluated the results.
• Spot gaps and bias. Encourage students to compare AI-generated responses with human-written sources to find missing perspectives or incorrect assumptions.
• Know when not to use AI. Help students reflect on when AI is useful—and when it’s better to rely on their own understanding. If AI decides what’s important, what learning might be missed?
• Do the task, not the prompt. Sometimes the most efficient approach is doing the task directly. Encourage students to consider when it’s worth prompting an AI and when it’s better to think or write it through themselves.
• Practice and talk about use. Provide students with opportunities to try Gen AI, review results with peers, and reflect on its use. Discussion is vital for supporting students to be active, critical users of Gen AI.

Professor Examples

Varied Options for Gen AI Use

Students need opportunities to test and find the limits of Gen AI. A web site development professor had noticed that her students over-relied on Gen AI when they lacked confidence in their own skills. She used a “red light, yellow light, and green light” system for her core assignments. Red light is for foundational practice in which students completed the assignment without Gen AI to build their skills and confidence. Yellow light was for a few sections of the assignment where AI assistance was allowed. (This was the most common scenario). Green light allowed students to see what they could accomplish with no AI restrictions. She tried to include one green light activity within each course that she teaches to encourage experimentation.

Impact: By introducing in-class red light activities early in the course, students were able to develop key foundational skills and learning confidence through their own effort. The red, yellow, and green system was a means of scaffolding intentional student collaboration with Gen AI while building learner self-confidence.

Role of AI in their Discipline

In their finance course, students explore the role of AI in their profession. In groups, they imagine and conduct research on how AI may influence their future career. They listen to an industry guest speaker who discusses how they use AI in their work and the ways this may change over time. As an assignment, students prepare a summary report with an action plan in which they reflect on the mix of skills (foundational and AI) they need to develop for career success.

Impact: Early reflection on what effective Gen AI use looks like in their industry can reinforce the need to develop foundational skills without overreliance on Gen AI. Typically, proficient Gen AI use in a field requires developing foundational expertise.

Multi-part Assignment

A communications professor used a multi-part assignment to support students to complete a simple research and writing assignment. Structured over 7 weeks, the assignment required students to prepare a fictional vignette of a parental response to their child’s psychological disorder using Gen AI.

• Part one asked students to become familiar with three psychological disorders from two sources provided.
• Part two, which was done in class required students to complete a concept map outline of their vignette.
• Part three, which was also done in class, required students to collaborate with Gen AI to prompt it for examples of child behaviours, characteristics, and interests that could be added to the vignette. As they complete this, the professor reminded students of how Gen AI tools functions via prediction rather than true reasoning. Students were encouraged to look for biases, errors, and gaps in nuance.
• Part 4 required students to prepare a final version of the vignette.
• Part 5 required them to prepare a short audio or video reflective analysis of the experience with using Gen AI to prepare the vignette. In particular, they were asked to consider whether Gen AI assisted or hindered their understanding of a parent’s perspective of raising children with psychological disorders.

As part of their submission, student included a copy of the Gen AI prompt interaction. The reflection required them to refer to the interaction.

Impact: The assignment was successful in supporting students to work intentionally with GenAI and reflect critically on the process. Students reported that they enjoyed the opportunity to work with AI without the worry of being penalized. Most expressed ethical concerns they had about using Gen AI and several mentioned the need to fact check and think critically when using Gen AI.

Explore Further

AI as an Educational Ally: Innovative Strategies for Classroom Integration
This article by Lisa Delgado Brown describes how she went “meta” with AI by having students connect its use to broader ethical concerns. *You will need to make an account in order to see this article.

AI can be responsibly integrated into classrooms by answering the ‘why’ and ‘when’
This article by Soroush Sabbaghan, from the University of Calgary, uses a philosophical lens to emphasize that supporting students to understand why and when to use Gen AI is more important than focusing on the technical aspects of how to use the tools.

Reflection