Generative AI is changing how students learn, write, and solve problems, whether or not it’s part of your course. Some faculty prohibit AI, some allow it for specific tasks, and some build it into assignments. Each can be the right choice, depending on your discipline and your learning outcomes. This page helps you make a thoughtful decision, communicate it clearly, and design your course with AI in mind.
How AI Affects Learning
Research shows that AI can help or hurt learning, depending on how students use it.
- In a study of nearly 1,000 students, those who practiced math with an unrestricted AI chatbot did better during practice. Once the tool was taken away, though, they scored lower than students who never had it. A version designed to give hints instead of answers largely avoided that drop (Bastani et al., 2025).
- In a randomized study in a large physics course, students who used an AI tutor designed around good teaching practices learned more than students in an in-class active learning session. The tutor guided students step by step and didn’t give away answers (Kestin et al., 2025).
The lesson for teaching: AI that does students’ thinking for them can weaken learning, while AI designed to prompt students’ thinking can support it. Either way, the principles of good teaching still apply. See Evidence-Based Teaching Practices.
Decide How AI Fits Your Course
Start with your learning outcomes, not with the technology. Ask yourself:
- What skills must students build and show on their own?
- Where, if anywhere, could AI support learning, such as brainstorming, practice, or feedback?
- Where would AI replace the thinking students need to do?
- Will students need to use AI in their field or career?
Your choice can be one of these, and it can differ by assignment:
| Approach | What it means | When it fits |
|---|---|---|
| Prohibit | Students may not use generative AI | Students need to build foundational skills on their own, or the course outcomes depend on independent work |
| Limit | Students may use AI for specific tasks you identify | AI can support part of the process, such as brainstorming or feedback, but not the final work |
| Integrate | Students use AI as part of the assignment | Using and evaluating AI is part of what students need to learn |
Your syllabus includes a generative AI statement for each approach. See Syllabus Design.
Secure and Open Assessments
A practical way to plan is to think about two kinds of assessment in your course:
| Assessment Feature | Secure assessments | Open assessments |
|---|---|---|
| Purpose | Confirm what students can do on their own | Help students practice, get feedback, and learn |
| AI use | You control whether AI is allowed | AI use is possible, so set clear expectations and guide it |
| Examples | In-class exams, oral explanations, presentations with questions, and proctored online exams | Homework, drafts, discussions, projects, and practice quizzes |
| Assessment type | Mostly assessment of learning | Mostly assessment for and as learning |
Most courses need both. Use secure assessments for the outcomes students must show independently, and use open assessments to build the skills they’ll need to succeed on them. For more on the three purposes of assessment, see Assessment, Grading, and Rubrics.
Talk with Students About AI
Whatever you decide, students need to hear it clearly. They take courses with very different AI expectations.
- Explain your reasons. Tell students how your AI expectations connect to what you want them to learn. Students are more likely to follow expectations they understand.
- Be specific. State your expectations in each assignment’s directions, not only in the syllabus.
- Give examples. Describe what is and isn’t allowed in your course.
- Revisit the conversation. Remind students of expectations before major assignments, not only in the first week.
Design Assignments That Require Students’ Own Thinking
If you don’t allow AI, or want certain assignments to show students’ independent work, design those assignments so their own thinking is visible and necessary. No assignment is completely AI-proof, but these strategies make AI much less useful and make students’ own work easier to see.
Strategies That Work
| Strategy | Example | Why it works |
|---|---|---|
| Make the process visible | Break a paper into a proposal, annotated sources, an outline, a draft, and a final version, each submitted in Canvas | You see the work develop over time, and students get feedback along the way |
| Complete work in class | In-class writing, handwritten responses, problem solving at the board, or a two-stage exam | Students show what they can do in real time |
| Add an oral component | A short conference or presentation where students explain their work and answer questions | Students must explain and defend their thinking on the spot |
| Use specific course context | Ask students to apply a concept to a class discussion, guest speaker, lab result, or case you analyzed together | AI tools don’t know what happened in your class |
| Connect to students’ experience | Ask students to analyze their own leadership experience, internship, interview, or fieldwork | The work depends on information only the student has |
| Use original materials | Provide a unique data set, image, document, or scenario for students to analyze | Students work from material AI hasn’t seen |
| Ask for reflection | Have students explain their choices, what they struggled with, and what they would do differently | Reflection reveals understanding and builds self-awareness |
Test Your Assignment
Before you assign a task, try it yourself:
- Paste your assignment prompt into a generative AI tool.
- Review the response. Would it earn a good grade with your rubric?
- If it would, revise the assignment using the strategies above. For example, add a process step, a course-specific requirement, or an oral explanation.
Don’t enter student work or names when you test.
Grade What Matters
- Reward process. Give points for drafts, checkpoints, and revisions, not only the final product.
- Reward specifics. Include rubric criteria for connecting to course materials, class discussions, or personal experience.
- Reward reasoning. Include criteria for explaining and justifying choices.
Online Courses
- Use LockDown Browser for any online quiz or exam where AI isn’t allowed.
- Use video replies or short recorded explanations so students explain their thinking in their own voice.
- Hold brief one-on-one Zoom check-ins on major projects.
- Design discussion prompts that require students to respond to specific points from classmates or course materials.
Still Talk About AI
Students will use AI in other courses and in their careers. A short conversation about why you’ve made your choice, and what AI gets wrong in your field, builds understanding and trust.
If You Limit or Integrate AI
Set Clear Boundaries
- Identify which tasks AI may support, such as brainstorming, outlining, practice, or feedback on a draft.
- Explain how students should acknowledge or cite their AI use.
- Add a rubric criterion for how well students use, evaluate, or explain their AI use.
Assignments That Use AI on Purpose
- Critique the output. Students ask an AI tool a course question, then fact-check and evaluate the answer using course sources.
- Compare and improve. Students compare an AI-generated response with expert sources and explain the differences.
- Use AI as a study partner. Students ask an AI tool to quiz them on course content, then reflect on what they got wrong and why.
- Revise with reasoning. Students revise an AI draft and explain each change they made and why.
Tools That Support Your Approach
The Citadel provides tools in Canvas to support academic integrity, whatever approach you choose.
- Turnitin: Checks submitted work against other sources and creates a similarity report. Use reports as a starting point for a conversation with the student, not as proof on their own. Review the report, look at the student’s process and earlier work, and talk with the student about how they completed the assignment.
- LockDown Browser: Prevents students from opening other websites, applications, or AI tools during an online quiz or exam in Canvas. Use it for every online quiz or exam where AI and other resources aren’t allowed.
Tools work best alongside good assignment design and clear expectations. If you have concerns about a student’s work, follow The Citadel’s academic integrity process. For more, see the Academic Integrity Certificate.
Build Students’ AI Literacy
AI literacy is the set of skills people need to critically evaluate AI tools, work effectively with them, and use them responsibly (Long & Magerko, 2020). You can build these skills in any course, even one that prohibits AI, by helping students:
- Understand what AI can and can’t do. AI tools generate likely-sounding text and images. They can be wrong, make up sources, or leave out important context.
- Evaluate what AI produces. Check facts, sources, and reasoning against reliable sources.
- Use AI responsibly. Follow each course’s expectations, acknowledge AI use, and protect personal and sensitive information.
- Learn how AI is used in your field. Discuss how professionals in your discipline use AI, and where human judgment is still essential.
If You Use AI in Your Own Work
Some faculty use AI to help draft course materials, practice questions, rubrics, or announcements. If you do:
- Review everything. Check AI-generated content for accuracy and alignment with your outcomes before students see it.
- Keep your voice. Personalize anything you share, especially feedback and announcements.
- Protect student information. Don’t enter student work or personal information into AI tools unless The Citadel has approved the tool for that use.
- Keep it accessible. AI-generated images, documents, slides, and videos must meet the same accessibility standards as anything else in your course. Always review AI-drafted alt text and captions for accuracy. See Accessible Canvas Content.
Training and Support
- AI training for faculty: CEITL offers training to help faculty teach effectively with AI, or without it. This includes the Adapting Classes and Assignments for the AI Era course in the Advanced Online Teaching Certificate. Contact CEITL to learn about current offerings.
- Turnitin and LockDown Browser: CEITL can help you set up and use both tools in Canvas.
- Consultations: Schedule a course design and teaching consultation for help redesigning an assignment, whatever approach you choose.
- Campus AI resources: Visit AI for Teaching and Learning for The Citadel’s AI policies, guidelines, and events.
Sources
Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, Ö., & Mariman, R. (2025). Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences, 122(26), e2422633122. https://doi.org/10.1073/pnas.2422633122
Kestin, G., Miller, K., Klales, A., Milbourne, T., & Ponti, G. (2025). AI tutoring outperforms in-class active learning: An RCT introducing a novel research-based design in an authentic educational setting. Scientific Reports, 15, 17458. https://doi.org/10.1038/s41598-025-97652-6
Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design considerations. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems. Association for Computing Machinery. https://doi.org/10.1145/3313831.3376727
Have a question? Contact CEITL by phone, email, or in person in Bond Hall, Room 363.