A student asks, 'Can I use AI to help with this assignment?' You know what you intended, but the syllabus doesn't quite say it. Another student has already made a different assumption. That's a hard place to begin a fair conversation.
Yes, your syllabus should explain your expectations for generative AI when its use could affect coursework. The useful question is what students may do, on which tasks, and what they need to disclose. A clear example can prevent more confusion than a broad warning.
The Faculty Role in Setting Expectations:
Start with the institutional policies and course requirements that apply to you. Then connect the course expectations to your learning goals. If students need to demonstrate independent analysis, explain which work must be their own. If evaluating an AI response is part of the assignment, explain that task just as clearly.
Discuss the policy in class and include assignment-level instructions where needed. Clarify whether brainstorming, translation, coding help, grammar suggestions, or generated text are treated differently. Students shouldn't have to guess which uses fall under the word 'AI.'
Three Approaches to AI Policy Statements:
The examples below are starting points to adapt within your institution's requirements. Add the course details and permitted tools before using them.
1. No Generative AI Use on Designated Work:
Sample statement: 'The assignments identified as independent work in this course must be completed without generative AI assistance. This includes using AI to draft responses, generate an outline, solve problems, or write code for those assignments. The purpose is to practice and demonstrate [specific skills]. The instructions will identify any exceptions. If you're unsure whether a tool or use is allowed, ask before submitting. Approved accessibility arrangements will be addressed through the appropriate institutional process.'
2. Limited Use with Disclosure:
Sample statement: 'You may use generative AI for [permitted activities] on [specified assignments]. You may not use it for [restricted activities]. When you use it, include a brief disclosure naming the tool, explaining what you asked it to do, and describing how you checked and revised its output. You're responsible for the accuracy, sources, and final content of the work you submit.'
3. Guided AI Use as Part of Coursework:
Sample statement: 'Some assignments in this course ask you to use or evaluate generative AI. Follow each assignment's instructions about the tool, task, and documentation. Distinguish your own reasoning from generated material, check factual claims and sources, and explain how you evaluated the output. Where access is required, the assignment will explain the approved access route or an appropriate alternative.'
Why Clear Policies Matter:
- Clarity: A student should be able to tell whether a specific use is allowed before beginning the work. Include a permitted example and a restricted one.
- Responsibility: Explain what students must verify and disclose. A generated citation, calculation, or summary still needs checking before it becomes part of a submission.
- Fairness: Consider access, cost, and required accounts when designing an AI-based task. Don't assume every student has the same tools available.
- Privacy: Tell students not to enter private information about themselves or others, confidential materials, or restricted course content into a tool unless that use is explicitly authorized through the applicable process.
- Consistency: Connect the syllabus, assignment instructions, and existing academic integrity procedures. Explain how questions or concerns will be addressed rather than inventing a separate process in an individual assignment.
A policy becomes useful when students can apply it. Try a short scenario in class: a student asks AI to generate an outline, checks the sources, and writes the paper. Is that allowed here? What disclosure would be needed? Talk through the answer using your actual instructions.
Before the next assignment, write down one allowed use, one restricted use, and the disclosure you expect. If those details are hard to state, the policy probably needs another pass. Give students a clear route to ask before uncertainty turns into a dispute.
