If you are an instructor you’ve likely written some version of an Artificial Intelligence (AI) policy into your syllabi. If you’re an instructional designer, you’ve probably read dozens of versions of these policies or might have helped craft one. An AI policy is an important first step in communicating to our learners what is allowed, what is not, and how AI could impact their learning. In fact, the faculty senate at Oregon State University recently voted that an AI policy on course syllabi is now required that explicitly communicates expectations for student AI use. This is a step in the right direction, but a policy alone is not enough.

As we already know, many learners are actively using AI tools. They are forming opinions, building habits, and wrestling with hard questions about what responsible and ethical use looks like. The real question is not whether students will engage with generative AI, but whether or not we will engage with them about it.

What Students are Saying

In 2024 the OSU Ecampus Research Unit (ECRU) surveyed online students about their perceptions and use of Generative AI (GenAI). Their findings are worth thinking about more closely.

In this study, students shared that they have a good working knowledge of GenAI tools, and they also voiced skepticism about the information those tools produce. On the other hand, they also believed that AI would shape their future careers. Additionally, many described a frustrating lack of clarity related to AI course policies. They feel unsure from one class to the next what is actually expected of them. OSU learners voiced more concerns than hopes about the role of AI in their education. These findings indicate that we have work to do when it comes to clarifying the role of AI in our learning spaces.

This confusion is not unique to OSU either. A 2026 report from MIT described how inconsistent guidance leaves students anxious and uncertain about when and why AI is welcome. The report uses the term: “cognitive surrender,” which is the tendency to reach for AI at the very first hint of struggle. You might also hear this called “cognitive offloading” or “cognitive outsourcing.” No matter the term, the basic idea is the same, students who turn to AI first are missing out on the productive struggle that is inherent in the learning process. When we remember and remind students that struggle is where much of the real learning happens, the stakes of this moment come into focus.

A Policy is Just the Beginning

This is where it gets interesting for those of us who design learning experiences. The studies indicate that students want (and need) more from us than a blanket policy, they want to know why and when and how. Simply stating “AI Use Prohibited,” is no longer enough.

Instead, consider sharing the skills that the assignment is meant to build, and the reason why having learners do it themselves matters. This small shift, from declaration to explanation changes the whole conversation. It treats students as partners in their own learning, rather than suspects to be monitored.

Relatedly, a 2025 faculty survey conducted by ECRU found that 32% of instructors admitted to never disclosing their own AI use to students, while at the same time many are requiring AI use-disclosure from students. Students notice that kind of asymmetry. If we want to build trust and model responsible use, transparency has to run in both directions. My colleague Naomi Aguiar covers this more fully in another recent post: “AI Disclosure Matters: Closing the alignment gap between online instructors and students.”

Building it Together

So, what does it look like to bring students into the conversation? It means creating a genuine space for dialogue. Here are a few places to start:

• Share early and often. A syllabus policy is a good start, but adding specific AI Guidance on each assignment is the next step.
• Share the why. When you draft AI Guidance for an assignment, explain the learning it is meant to protect or develop. This rationale invites buy-in in a way a simple stated rule never will.
• Invite the discussion early. Consider using an introductory discussion to ask students how they are already using AI, what worries them, and where they see it helping in this course. Engage with them without judgement, sharing your own concerns and examples.
• Co-create your norms. Consider building a shared set of expectations for responsible use, so students have real ownership of the guidelines they are asked to follow.
• Model transparency. Tell students when and how you use AI in the course, and why. Your example carries more weight than any policy ever could
• Keep the human at the center. Your presence in discussions and your feedback build a connection no tool can replicate. Students don’t want AI-generated responses from you or their peers. It’s okay to name that, and to share how and why you show up for them personally.

These simple steps do more than clarify the “rules.” They build a shared understanding of what we are here to do together.

Why it Matters

We are not just teaching our content in higher ed; we are teaching students how to think critically, how to engage in struggle as a learning practice, and how to evaluate resources. More importantly, we are preparing them for life beyond the classroom. That is the work in front of us. Our learners already live in a world saturated with these tools, and we can help them build healthy habits of the mind: curiosity, skepticism, ethical judgment, and the confidence to think for themselves. These habits are cultivated, but only if we first plant the seed, and continue to nourish it. For our online students, this starts with how courses and activities are designed, the guidelines and resources provided to them, and the way we show up alongside them.

Stating our expectations is a good starting point. Building understanding together is how we move this work forward and show our commitment to student learning and success. Let’s make room for the conversation.

Citations:

Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training. (2026). Report of MIT’s Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training. Massachusetts Institute of Technology. https://aiandeducation.mit.edu/report/

Aguiar, N. R., Coltharp, A., Dello Stritto, M. E., Jorgensen, J., Kronser, Z. E., Pearson, A., Piacenza, S., & Reese, D. (2026). Online Faculty Perceptions of Generative AI. Oregon State University Ecampus Research Unit. https://ecampus.oregonstate.edu/ research/publications/

Dello Stritto, M. E., Underhill G. R., & Aguiar, N. R. (2024). Online Students’ Perceptions of Generative AI. Oregon State University Ecampus Research Unit. https://ecampus.oregonstate.edu/research/publications/