Source-linked AI summary
A Guided Inquiry Approach to Students Co-Designing Generative AI Course Policies
Ashish Hingle, Aditya Johri
TL;DR
As GenAI use expands amid inconsistent educational policies, this exploratory study examines a guided inquiry activity in which students co-design GenAI course policies. Students developed individual proposals, refined them with peers and stakeholder perspectives, and reflected on the process; they emphasized preparation, disclosure procedures, institutional support, and participation in decision-making.
Problem
Inconsistent institutional GenAI policies and students’ active engagement with AI-use questions motivate examining how students conceptualize and construct course policies.
Method
Students used guided inquiry to research GenAI, create individual policy proposals, collaborate with peers on a collective policy, and reflect on the design process.
Results
Participants prioritized student and instructor training, standardized AI-use disclosure procedures, stronger institutional support, and greater involvement in GenAI decision-making.
Takeaways & Limitations
Policy co-design can surface student values, concerns, multiple perspectives, and trade-offs involved in governing GenAI use in education.
Takeaways & Limitations
The exploratory activity has a preliminary, context-bound scope, so its findings are context-specific.
Abstract
from arXiv · showhide
As generative AI (GenAI) use among students increases, educators face growing questions about how to support learning while addressing ethical and institutional concerns. This exploratory study examines a guided inquiry activity in which students co-designed a GenAI course policy. Students first developed individual policy proposals focused on appropriate and ethical use of GenAI, then collaboratively refined them by incorporating diverse stakeholder perspectives. The following research questions guided the study: 1) what practical factors do students prioritize in their GenAI use policies, and how do they justify these choices? and 2) how do participants reflect on the policy design process? Participants first completed readings, then used GenAI to brainstorm initial policy ideas. Next, they articulated their own perspectives through a written assignment and a course policy they designed individually. Finally, they incorporated diverse stakeholder perspectives by collaborating with peers to develop a collective policy. Analysis of student artifacts and group discussions showed that participants prioritized training for students and instructors, standardized procedures for disclosing AI use, and stronger institutional support. Participants also wanted greater involvement in GenAI-related decision-making. They described the policy design process as a way to engage with multiple perspectives and the inherent trade-offs involved in governing AI use. This study offers pedagogical insights into how policy co-design activities can surface student values, concerns, and sensemaking about GenAI in educational contexts.
I. Introduction
The study addresses inconsistent institutional GenAI policies by examining how students conceptualize and construct course policies through guided inquiry. It asks which practical factors students prioritize and how they reflect on the design process.
- Institutional approaches to acceptable GenAI use and enforcement remain inconsistent across U.S. higher education.
- Students are already engaged with questions about what AI is and how it should be used in learning.
- The study presents a guided inquiry activity in which students design GenAI course policies while researching independently and navigating perspectives with peers.
- The research examines which practical factors students prioritize in proposed policies, how they justify those choices, and how they reflect on policy design.
II. Literature Review and Conceptual Model
The conceptual model treats GenAI policy as both a student-engagement artifact and a pedagogical tool, combining guided inquiry, perspectival thinking, student voice, and AI literacy. Students are positioned as co-designers who consider multiple stakeholders and negotiate educational values.
- The study reframes GenAI policy as a social process through which students negotiate acceptable use, responsibility, and academic values.
- Conceptual Model: The conceptual model positions students as co-designers of course structures and policies that affect them.
- Guided Inquiry Approach: Guided inquiry supports student-centered exploration, critical questioning, and scaffolded problem-solving in GenAI education.
- Engaging Perspectival Thinking and Student Voice: Perspectival thinking asks learners to consider multiple stakeholders when designing course policies.
- Engaging Perspectival Thinking and Student Voice: Co-design positions learners as contributors to the educational environments, tools, and policies shaping their experiences.
- AI literacy is framed as context-specific, spanning technical proficiency, critical understanding, and ethical awareness.
III. Study Context
The study was embedded in an undergraduate technology-ethics course module on GenAI in education. Students progressed from curated resources and individual policy proposals to peer discussion and collective policy development.
- The study took place in an undergraduate technology-ethics course examining information technology in societal and systemic contexts.
- The GenAI module taught how GenAI tools function, effective prompting, and response-accuracy evaluation before the culminating group discussion.
- Resources: Students received curated resources on GenAI’s technical and societal dimensions and were encouraged to extend them through additional research.
- Individual Course Policy Design Assignment: Each student proposed at least five policy elements, explaining their importance, benefits, and possible concerns for other stakeholders.
- Individual Course Policy Design Assignment: Students drew on provided resources, independent research, and personal experiences of GenAI’s learning benefits and challenges.
- Group Discussion & Writing Assignment: Groups of 4-6 peers spent approximately 45 minutes presenting views, questioning proposals, considering implementation, and producing a final GenAI policy.
- Group Discussion & Writing Assignment: Participants reflected on changes between individual and collaborative policy creation through an immediate debrief and a same-day written assignment.
IV. Methodology
The exploratory study analyzed consenting undergraduate students’ policy assignments and group discussions using hybrid inductive-deductive thematic analysis. Coding was organized with the ED-AI Lit Framework and reached Cohen’s Kappa (.84) reliability, with procedures intended to reduce grading-related bias.
- Participants: 68 undergraduate students enrolled in the course, and 55 consented to participate in the study.
- Participants: Participants were mostly junior and senior information technology and cybersecurity students with varied prior AI and GenAI experience.
- Data Analysis: The research team analyzed assignments and group discussions using a hybrid inductive-deductive thematic analysis approach.
- Data Analysis: Cohen’s Kappa (.84) measured interrater reliability, and coding discrepancies were resolved through discussion until reviewers agreed.
- A non-teaching researcher led the initial data review, and grades were posted before analysis to reduce potential bias.
V. Findings
Participants framed effective GenAI course policy as an interconnected system spanning knowledge, evaluation, collaboration, contextualization, autonomy, ethics, and institutional responsibility. Their proposals emphasized training, transparent disclosure, supportive accountability, equitable access, approved tools, privacy, and ongoing revision.
- Findings: Participants organized policy design around seven interconnected themes, with institutional responsibility forming a final macro-level theme.The analysis used the ED-AI Lit Framework and examined participants’ reasoning and intentions behind proposed policy elements.
- Knowledge and Evaluation: Students prioritized task-related training for both students and instructors to support informed and context-specific GenAI use.They emphasized accessible training, awareness of limitations, critical evaluation, prompting, and instructor confidence.
- Collaboration and Contextualization: Participants requested standardized disclosure and citation procedures while favoring supportive systems over punitive responses to mistakes and misuse.They also distinguished encouraged uses from prohibited activities and supported course-specific clarification of permissible AI use.
- Autonomy, Ethics, and Institutional Responsibility: Students emphasized accountability, equitable access, environmental sustainability, and institutionally approved tools as components of responsible GenAI policy.Proposals also addressed data collection, privacy, security, and the need to balance convenience with institutional obligations.
- Institutional Responsibility: Participants called for policies to be regularly reviewed and updated as GenAI capabilities evolve.They also viewed student involvement in policy development and revision as important for trust among students, instructors, and institutions.
i. Theme 1: Knowledge
Participants treated knowledge, evaluation, collaboration, and contextualization as connected foundations for responsible GenAI use. They sought training, critical evaluation, practical prompting, disclosure, and course-specific guidance rather than isolated rules.
- Theme 1: Knowledge: Participants viewed knowledge of GenAI capabilities, limitations, and risks as foundational to interpreting and applying policy consistently.They called for accessible, task-related training customized to students’ fields or courses, alongside instructor training.
- Theme 2: Evaluation: Participants emphasized critical evaluation of AI outputs to address inaccuracies, bias, training-data limitations, and potential misuse.They differed on whether policy should restrict AI broadly or promote critical engagement with its limitations and biases.
- Theme 3: Collaboration: Participants recommended prompting exercises and sample prompts to help students use GenAI productively, efficiently, and transparently.Prompting was framed as a teachable practice connected to disciplinary learning rather than a technical shortcut.
- Theme 4: Contextualization: Participants argued that appropriate GenAI use depends on discipline, assignment type, and learning goal.They supported explicit guidance about permitted uses and policies tailored to specific courses rather than one universal rule.
v. Theme 5: Autonomy
Participants linked student autonomy to knowledge, evaluation skills, contextual guidance, and responsibility for AI-assisted work. Their reflections and policy themes also foregrounded fairness, access, privacy, sustainability, institutional responsibility, and ongoing participation in policy decisions.
- Theme 5: Autonomy: Participants defined autonomy as informed self-determination supported by GenAI knowledge, evaluation skills, and contextual guidance.They also stressed that students remain responsible for AI-assisted work and its consequences.
- Theme 6: Ethics: Participants connected responsible AI use with fairness, accountability, transparency, privacy, and equitable access to tools.They viewed ethical use as dependent on both individual behavior and broader social and institutional conditions.
- Theme 6: Ethics: Participants argued that environmental impact should shape whether GenAI use is necessary, equitable, and educationally justified.Several advocated discouraging uses that do not meaningfully contribute to learning and reducing unnecessary queries.
- Theme 7: Institutional Responsibility: Participants assigned institutions responsibility for fair, consistent, secure GenAI use through approved tools, privacy protections, and regular policy revision.They also wanted students to have a say in developing and revising policies to build trust.
- Participant Reflections: About two-thirds of participants reported engaging with perspectives beyond their own, including peers, instructors, institutions, and employers.Approximately half reported a personal perspective shift, often involving greater awareness of GenAI trade-offs and instructors’ decision-making.
VI. Discussion and Implications
The exploratory, context-bound study suggests that guided inquiry and policy co-design can surface students’ perspectives on GenAI governance, while highlighting the need for scaffolding, instructor and institutional support, and further research on behavioral and cross-disciplinary outcomes.
- Student participation: Students wanted meaningful involvement in shaping GenAI course policy but remained uncertain about how to contribute effectively.The authors connect this tension to the distinction between consultation and substantive participation.
- Pedagogical implications: The guided inquiry activity supported exploratory pathways, perspectival thinking, and awareness of the trade-offs involved in GenAI policy design.Some students pursued prior assumptions, while others reaffirmed their interest in exploring AI tools further despite feeling unprepared.
- Instructional and institutional support: Participants linked student preparedness for AI-enhanced learning to instructors’ AI literacy and broader institutional learning efforts.They also raised questions about responsibility, authority, and communication among instructors, administrators, and students.
- Pedagogical implications: The activity can be adapted across disciplines to support AI literacy through policy topics connected to professional practice and future work.The authors present policies, rules, and standards as design topics that can extend beyond technical fields.
- Limitations and future research: Future research could examine changes in actual AI-use behavior, cross-disciplinary shifts in policy priorities, and translation of student-led discussions into institutional decision-making.Suggested measures include reflective logs, disclosure statements, surveys, and analysis of students’ evaluation and revision of AI-generated outputs.
- Limitations and future research: Because the activity was tailored to one class and lacked a baseline or preintervention measure, its findings describe articulated perspectives rather than attributable learning gains.The activity’s dual role as a graded assignment and research data source may also have influenced student responses.