Co-Creating a GenAI Class Policy: A Collaborative Inquiry
The Approach
Rather than arriving to the first class with a pre-written GenAI policy, Professor Moghaddam designed a structured in class activity for her Winter 2026 course, LINC10 (Argumentation and Analysis), where students were given both the responsibility and the tools to build the policy themselves.
The activity is built around a student handout organized into four required sections: Allowed Uses, Not Allowed, Gray Areas, and Justification. The activity unfolded across three phases.
Learn more in Professor Moghaddam’s student handout.
Phase 1: The Warm-Up Provocation
Students were first asked to think about a simple but intentionally unclear question: Would this count as plagiarism? They were given four scenarios, ranging from using AI to fix grammar to using it to write a paragraph, and asked to think about where they personally drew the line. This warmup helped surface their assumptions, brought out differences in how they saw the issue, and created space for real discussion rather than simply handing them a policy to accept.
Phase 2: Small-Group Policy-Building
Working in groups of two or three, students were given a structured handout organized around four required sections:
- Allowed Uses: What is clearly acceptable?
- Not Allowed: What crosses the line?
- Gray Areas: What depends on context?
- Justification: Why?
The four-part framework was intentional. It moved students away from simply thinking in terms of yes or no and asked them to explain their reasoning, not just sort tools into categories. The Gray Areas section especially pushed them to think more carefully about context rather than treating everything as absolute.
Phase 3: Class Debrief and Policy Finalization
Groups shared their positions in a full-class discussion. Professor Moghaddam facilitated without imposing conclusions, and the class arrived at a shared policy through consensus. Key outcomes included:
- Permitted uses: revising outlines, revising essays, brainstorming
- Not permitted: having AI generate ideas, arguments, or claims
- Context-dependent: summarizing and translation – acceptable for supporting personal understanding, not acceptable when submitted for a grade
The debrief also brought out students’ anxiety around AI detection tools. Professor Moghaddam addressed this openly and directly, explaining that she does not rely on these tools because of their well-documented limitations, including her own experience of having her writing flagged as AI generated. Being transparent about this helped ease student anxiety in a meaningful way.
The class concluded by agreeing to include an AI Acknowledgment section in their final projects, modelling the kind of clear and transparent disclosure they had been discussing and evaluating through the handout.
Pedagogical Rationale
Building Confidence Through Agency and Transparency: Co-construction shifts the process away from simply handing students a policy from the top down and toward shared ownership. When students help shape the rules, they are more likely to trust them and follow them. In this class, that sense of ownership also clearly helped reduce anxiety around AI use.
Structured Ambiguity as a Teaching Tool: The Gray Areas category was pedagogically deliberate. Rather than a binary framework, it modelled the contextual thinking that ethical AI use requires, and it pushed students to argue for their positions, not just declare them.
Transferable AI Acknowledgment Practices: Looking at different versions of AI disclosure statements gave students a practical skill they could carry into other contexts. The class’s decision to include an AI Acknowledgment section in their final projects came out of that discussion naturally. It was something they chose themselves, not something imposed on them.
Student Feedback
Professor Moghaddam shares: “The students were very engaged and had a lot to say. One student said, ‘I have never had this before in any course, and it is a new concept, and I love it!’ They mentioned that the fear around AI use was kind of gone, and they felt more confident because they had a role in shaping the policies.
What was most useful was that students had to explain their thinking. It moved the discussion away from a simple ‘AI is good’ or ‘AI is bad’ approach. They started thinking more carefully about learning, ownership, fairness, and what it really means for work to be their own.”