Small Bytes Blog

Beyond the Syllabus Statement: How to Actually Talk to Students About AI

By Suzanne Tapp, Associate Vice Provost of Teaching and Learning

The fall semester is underway, and whether you included an AI statement in your syllabus or not, one thing is clear: a few lines of text are not enough. Some faculty came into the semester with carefully crafted AI policies but others may still be questioning if their policies make a difference. Either way, the more important question right now is the same: how do you actually talk to your students about AI in a way that builds understanding, sets clear expectations, and creates the kind of classroom culture where good decisions get made? The research and practice coming out of universities around the world offer some genuinely useful guidance.

So what does a syllabus statement actually do?

Dr. Tazin Daniel at the Center for Research on Learning and Teaching (CRLT) at the University of Michigan makes a point worth sitting with before you write a single word: the challenge with AI syllabus statements is rarely finding the perfect wording. It is expecting too much from a few paragraphs in the first place.

Instead of asking “how do I write the perfect AI statement,” she suggests starting with a different question: “what can I reasonably expect this statement to do?”

A well-crafted syllabus statement can do four things well. It can clarify your expectations for when and how AI may or may not be used in your course. It can explain the learning rationale behind those expectations, so students understand not just the rule but the reason. It can signal your course values, communicating through its tone whether your classroom is built on trust and curiosity or surveillance and suspicion. And it can invite ongoing dialogue, making clear that this is a conversation that will continue throughout the semester rather than a policy handed down on day one.

What it cannot do, no matter how carefully you write it, is prevent academic misconduct, teach AI literacy, replace thoughtful assignment design, or anticipate every possible scenario. When we expect a syllabus statement to carry all of that weight, it becomes overloaded and less effective.

Move beyond the statement

A syllabus statement is a starting point and we want to emphasize that we hope you have this on your syllabus. Monash University’s teaching and learning team has developed one of the most practical frameworks for thinking about what comes next, organized around five actions: define, explain, bridge understanding, discuss, and educate.

Define means setting clear, specific expectations for each assignment rather than relying on a single course-level statement. Students benefit from knowing not just that AI use is “allowed” or “not allowed” but exactly what that means for the task in front of them. Texas Tech University’s AI Use Levels offer a practical tool for doing exactly this. The four levels move from Level 0, which requires students to complete all intellectual work independently without any AI assistance to Level 1, which permits AI for learning support and planning such as brainstorming or generating practice questions. Level 2 allows AI for specific purposes like feedback, revision, or coding support while requiring students to originate the central ideas and own the final product, and finally Level 3 says that AI is integrated as part of the learning outcome itself and evaluates students not just on the final product but on the quality of their prompts, their verification process, and their ability to distinguish their own contributions from AI. Communicating which level applies to each assignment removes ambiguity and helps students make informed choices rather than guessing at what you expect. It seems like another starting point to help them consider how they’ll use AI after graduation too.

Explain means helping students understand the difference between AI tools and contract cheating sites, which are fundamentally different in their intent and design. It also means being honest about how AI tools actually work, which brings us to the third action.

Bridge understanding means giving students a basic working knowledge of what large language models are and are not. They are not primary sources. They are probability engines that generate likely outputs without genuinely understanding meaning. They reflect the biases present in their training data. They may produce content that sounds authoritative but is factually wrong. Students who understand this are far better equipped to use AI critically and responsibly than students who simply receive a rule about whether they can or cannot use it.

Discuss means keeping the conversation going. Monash emphasizes that AI literacy is not built through a single conversation or a required module. It is built through ongoing engagement, open questions, and willingness to revisit expectations as the semester unfolds and as the technology itself continues to evolve.

Educate means grounding the conversation about AI in academic integrity, not as a compliance exercise but as a values conversation. When students understand why integrity matters in the context of their discipline and their future profession, they are more likely to internalize those values than when they simply encounter a rule to follow.

Remember that trust is the foundation

All of the practical guidance above rests on something more fundamental. Professor Kelly Matthews of the University of Queensland, writing in the International Journal for Students as Partners, argues that the most important resource we have for navigating AI in higher education is not policy but trust: what she calls pedagogical trust, “a confident, reciprocal learning relationship between students and teachers that is open to uncertainty and co-navigated through dialogic decision-making and shared sense-making” (Matthews, 2025, p. 1). This goes well with the Pedagogy of Kindness that we’ve emphasized at the TLPDC for the past year or so. That idea resonates deeply with a framework we have been emphasizing at the TLPDC: the pedagogy of kindness. Cat Denial, in her book A Pedagogy of Kindness, reminds us that kindness in teaching is not about niceness. It is about two things: believing people and believing in people. Believing people means extending good faith when a student tells you something is hard or that life got in the way. Believing in people means seeing students as collaborators who bring real experiences and perspectives into your classroom. When you are clear about your AI expectations, you are believing in your students' ability to meet them. When you extend good faith rather than suspicion, you are practicing exactly the kind of pedagogical trust that Matthews describes. As our TLPDC director Dr. Cam Brown often says, clarity equals kindness. The two frameworks are not the same, but they point in the same direction: toward classrooms where students feel seen, trusted, and genuinely supported.

Matthews points out that AI has amplified deficit views of students, framing them as either cheaters or victims. That framing, she argues, risks damaging the very relationships that make learning possible. With 83% of students in a large multi-institutional study reporting that they use AI, shame and suspicion are not only unproductive; they are disconnected from the reality students are actually living.

Her five principles for navigating AI through a partnership lens are worth keeping close as you consider your communication strategy. Cultivate open and curious conversations where uncertainty is treated as an opportunity rather than a threat. Acknowledge the full range of emotions students and faculty bring to this topic, including anxiety, shame, confusion, and genuine curiosity. Create consistent opportunities for shared reflection as the semester progresses. And ground your expectations in shared values rather than surveillance.

The language you use to talk about AI with your students is itself a pedagogical act. It communicates whether you see them as partners in navigating something genuinely new, or as problems to be managed.

A practical starting place

If you are not sure where to begin, here are three things you can do right now:

Add or revisit your syllabus statement so that it clarifies your overall expectations, explains your reasoning, and makes clear that you welcome questions. It does not have to be perfect. It has to be honest and specific enough to give students something to work with.
Add assignment-level language for each major assessment that tells students exactly what AI use looks like for that particular task. Texas Tech’s AI Use Levels give you a ready-made framework for doing this consistently across your course.
Plan a class conversation about AI if you have not already. Not a lecture about policy, but an open discussion about how students are already using these tools, what questions they have, and how you will navigate the semester together. That conversation will do more for your classroom culture than any syllabus statement ever could.

The goal is not to write or include the perfect policy. It is to build the kind of classroom where students feel equipped, trusted, and genuinely supported to make good decisions about a technology that is not going away.

Disclaimer: I used generative AI (Claude) solely to improve grammar, clarity, and word choice in this manuscript. The research, arguments, and conclusions come from my caffeine-supplemented brain.

References

CRLT, University of Michigan. (n.d.). Making the most of your GenAI syllabus statement. https://crlt.substack.com/p/making-the-most-of-your-genai-syllabus

Denial, C. J. (2024). A pedagogy of kindness. University of Oklahoma Press.

Matthews, K. E. (2025). Five guiding principles for navigating artificial intelligence in students as partners practice to preserve pedagogical trust. International Journal for Students as Partners, 9(2), 1–8. https://doi.org/10.15173/ijsap.v9i2.6871

Monash University. (n.d.). Communication with students about AI. Monash University Teaching and Learning. https://www.monash.edu/learning-teaching/teachhq/Teaching-practices/artificial-intelligence/communication-with-students

Texas Tech University. (2026). Generative AI in teaching and learning: Guidance, recommendations and sample syllabus statements. https://www.depts.ttu.edu/tlpdc/AI-Syllabus-Statements.php

By Suzanne Tapp, Associate Vice Provost of Teaching and Learning

Learn more

AI Teaching Resources

AI Tools for Research & Writing | University Libraries

The AI Small Bytes is a blog written by members of the AI Resources & Guidelines Committee. This blog will be updated periodically with new resources and information, and we hope that you will check back often. If you would like to talk about teaching with artificial intelligence and your concerns or ideas, please feel free to contact Lisa Low, Director of the Division of AI Pedagogy and Policy.