Participatory, not Punitive: Student-Driven AI Policy Recommendations in a Design Classroom

Paper Title

Participatory, not Punitive: Student-Driven AI Policy Recommendations in a Design Classroom

Publication Info

• Topic area: Participatory design and governance of AI policies in higher education.
• Keywords: Generative AI, participatory governance, student-driven policies, design education, AI ethics, zine-making, higher education, academic integrity, AI literacy, equity.

Background and Problem

• Problem / challenge: University AI policies are predominantly top-down, punitive, and exclude student perspectives, leading to confusion, fear, and ineffective governance.
• Significance: Students are the primary stakeholders affected by AI policies, and their exclusion results in policies that fail to address their lived experiences and concerns, such as overreliance on AI, inequity, and inconsistent enforcement.
• Motivation and related work: Prior work highlights the need for participatory approaches in AI governance to improve trust, efficacy, and inclusivity. However, existing efforts often remain tokenistic, and there is little evidence of student input shaping AI policies in higher education.

Solution

• Proposed approach: A three-part, student-led workshop series to co-create AI policy recommendations in a graduate design course, culminating in a zine to disseminate the policies.
• Novelty:
  1. Empirical insights into students’ candid AI practices, including misuse and policy gray areas.
  2. A transferable model for student-driven AI policy design using faculty-free workshops and zine-based participatory infrastructuring.
  3. Ten actionable, student-authored AI policy recommendations addressing gaps in existing governance.
• Procedure and key techniques:
  1. Workshop 1: Candid discussions on AI use and initial policy drafting.
  2. Workshop 2: Zine-making to visualize and refine policies.
  3. Workshop 3: Applying policies through a design activity and reflecting on their practicality.
  4. Follow-up interviews and thematic analysis to validate findings and refine recommendations.

Results

• Concrete findings:
  • Ten student-authored AI policy recommendations, including guidelines for AI use, ownership thresholds (50% AI contribution), support for English learners, and equity in tool access.
  • Students reported increased intentionality in AI use, treating AI as a collaborative partner rather than a shortcut.
• Advantage over baselines:
  • Policies addressed gaps in top-down governance, such as assignment-specific guidance, faculty transparency, and nuanced support for English learners.
  • Participatory process fostered critical thinking, compliance, and improved student-instructor relationships.
• Experiments / evaluation:
  • Workshops with eight graduate design students at UMBC, a minority-serving institution.
  • Zine circulated across campus, sparking broader discussions among students, faculty, and administrators.
  • Follow-up interviews and expert evaluations of student-designed AI interfaces.
• Limitations and future work:
  • Limited to one course and university, reducing generalizability.
  • Policies may conflict with learning objectives or be challenging to implement due to enforcement and resource constraints.
  • Future work includes scaling participatory models across disciplines and addressing feasibility challenges.

Summary

This study demonstrates the value of engaging students in AI governance through a participatory, student-led process. By co-creating ten actionable policy recommendations and visualizing them in a zine, the project addressed gaps in top-down AI policies, such as inequity, faculty transparency, and support for English learners. The participatory process itself fostered critical thinking, compliance, and broader campus discourse, offering a transferable model for involving students in AI policy design. While challenges remain in implementation and alignment with learning objectives, the findings highlight the importance of calling students into governance rather than excluding them.