Teach with AI

As AI becomes more a part of our life, we need a simple way to show students where these tools help and where they hurt. A great way to do this is with a three-tiered framework for your assignments or your course as a whole. This eliminates confusion for students as they switch between different classes and assignments.

Here is how the three levels could work:

• No use: Students do the work completely on their own without any AI help. This is perfect for foundational steps where the struggle is the learning—like brainstorming an original thesis or writing a first draft by hand.
• Limited use: Students can use AI behind the scenes as a helpful partner for specific, restricted tasks. They might use it to build an outline, fix grammar, or quiz themselves, but the final writing and core ideas must still be their own, and they must disclose where and how they used AI.
• Extensive use: Students work directly alongside the AI with few or no restrictions on how the tool is used for that assignment. The goal here is usually to allow students to focus on other areas of learning, or to critique, edit, or fact-check what the machine wrote. This level does not negate a student properly disclosing and attributing the use of AI.

Example Syllabus Language

These examples can be the basis of the AI use section of your syllabus, and can be customized to fit your pedagogical approach.

Useful Links

• AI in your Syllabus: A presentation by Clark’s Mark Jacobs offering faculty a practical “Red, Yellow, Green” model and template language to clearly define permissible and prohibited AI uses on their syllabi. (External Site)
• Lance Eaton’s Syllabus Database: Presents a massive, crowd-sourced database of real-world syllabus policies from over 300 higher education institutions, offering faculty a valuable reference for how peers across various disciplines are framing AI guidelines. (External Site)

AI Peer Review: Students write, program, or create a first draft, ask an AI to critique their logic and evidence, and then revise it. They turn in both their prompt history, and a short reflection explaining why they chose to accept or reject the AI’s specific suggestions.

Fact-Checking: Give students an AI-generated paper containing fake citations or subtle errors. Have them use real library databases to track down the truth and turn in an audit report correcting the machine’s mistakes. Alternatively, give students AI generated code that produces the correct output, and have them identify inefficiencies, insecure variables, etc.

Socratic Dialog: Have students use a prompt that turns the AI into a Socratic tutor. Students must engage in a back-and-forth conversation to learn a concept and submit the chat transcript along with a paragraph summarizing what they learned. (Clark’s own Mark Jacobs has created a custom Socratic Dialog tool, External Site)

Role Plays: Have AI play the role of a client, historical figure, or examiner. Students use a role-play prompt to pitch a project, defend an idea, or test their understanding.

Useful Links

• AI for Education (AI Pedagogy Project, External Site)
• Socr-AI-tic Lab for Collaborative Pedagogy (Created by Clark’s Mark Jacobs, External Site)

Academic integrity remains the foundation of teaching and learning at Clark, regardless of whether AI tools are used. Faculty should help students understand that the responsible use of AI does not replace original thinking, disciplinary knowledge, or personal accountability. Students are responsible for the work they submit, including any content created or informed by AI, and AI use should never misrepresent authorship, learning, or intellectual contribution. Clark’s AI policy emphasizes transparency, appropriate attribution, and individual responsibility for AI-assisted work.

Faculty should use caution when interpreting results from AI-detection tools. These tools cannot reliably determine whether AI was used in the creation of a student’s work and should not be considered definitive evidence of academic misconduct. Instead, concerns about potential misuse of AI should be evaluated using academic judgment, course context, the nature of the assignment, and supporting evidence. Suspected violations should be addressed through Clark’s existing academic integrity process, which remains the appropriate mechanism for reviewing and resolving concerns about student conduct.

If you ask your class to use AI, you might be surprised to find some students strongly pushing back. This reluctance is often driven by deep ethical concerns, a fear of losing their own critical thinking skills, or a strict moral code regarding academic honesty. You don’t need to be an AI expert to handle these moments.

Validate Their Concerns: Don’t dismiss their objections. Acknowledge that their desire to protect their own voice, ethical boundaries, data privacy, or critical thinking skills shows a high level of academic integrity. Let them know you respect their perspective.

Explain the “Why” of the Tool: Clearly explain exactly why you are asking them to use it for this specific task. Emphasize that the goal isn’t to let the machine do the work, but to learn how to critique, edit, or audit AI outputs – skills they will likely need in their careers.

Keep the Focus on Student Judgment: Remind them that the AI is just a flawed baseline assistant. Reassure them that the real grade comes from their own human analysis, corrections, and original thought, keeping the heavy lifting completely in their hands.

Offer an Equal “Opt-Out” Path: If a student has a deep philosophical objection to using an AI tool, have a backup ready. Allow them to meet the exact same learning goals using traditional research methods or peer collaboration instead.

Be Honest About Your Own Limits: You don’t need to be an AI expert. If a student raises a complex question about data privacy or industry ethics that you don’t know the answer to, just say: “That is a massive question the field is wrestling with right now. Let’s figure out how to navigate this specific project safely together.”

Useful Links

• The Students Who Won’t Use AI (Teaching Upside Down, External Site)

Generate Classroom Analogies: If students are struggling with a highly abstract concept, ask an AI to brainstorm five different real-world analogies. This gives you quick, creative options to break down complex ideas in class.

Predict Student Misconceptions: Paste your upcoming lesson outline into an AI and ask it to identify the top three logical traps or points of confusion students usually face with that material. This helps you intentionally structure your lecture to fix those hurdles early.

Draft Rubrics and Feedback Templates: Feed your assignment prompt into an AI and ask it to draft a detailed scoring rubric based on your specific criteria. You can also use it to generate structured template responses for common student mistakes to speed up grading.

Create Low-Stakes Retrieval Quizzes: Drop your lecture notes or a course reading into an AI and instruct it to generate multiple-choice questions or short-answer prompts. It can instantly build quick, active-learning checks complete with an answer key.

Support Reading Materials: Paste a dense academic text or primary source into an AI and ask it to generate a companion guide. It can instantly create a vocabulary cheat sheet of specialized terms or a high-level summary to support diverse learning needs.

Design Role-Play and Debate Scenarios: Ask AI to generate role-play activities or detailed briefs for mock trials, case studies, or structured debates based on your course topics, providing ready-to-use small group activities.

Useful Links

• AI in Teaching (HBSP, External Site)
• Prompt Library for Teachers (AISchool, External Site)