Faculty Examples: Communicate with Your Students

Clarify your AI expectations with your students. Share your assignment-level AI policy, explain the reasoning behind it, and regularly invite clarifying questions.

This matters more than it might seem. As the AI and Learning page describes, silence can become its own barrier: students are often genuinely unclear about what's expected and don't ask, even when they want to. Saying your policy out loud, and explaining the reasoning behind it, closes that gap. See the Course & Assignment Design page for more on how this move fits into the fuller design picture. 

Below are examples from UO faculty.

Professor Durairajan redesigned CS 322, Introduction to Software Engineering, around a “Crawl, Walk, Run” progression. Students first learn to build on their own. Next, they use AI as a pair programmer while learning to critically evaluate its suggestions. Finally, they use AI more freely to build something ambitious. His goal isn’t simply to regulate AI use. It’s to help students develop the knowledge and judgment they need to use these tools well.

He explains: “I don’t want students to choose between learning software engineering and learning AI. They need both, but the order matters. First learn to think and build independently, then learn to work with AI, and finally use it to build something ambitious.”

He describes this approach in a recent paper, where one observation stands out. In end-of-term reflections, several students described a tension between using AI to get things done quickly and taking the time to understand what they were building. Some named specific concepts they skipped or didn’t retain because AI handled them. That tension is an important part of the motivation for the scaffolded approach. You can read more in his article, Crawl, Walk, Run: A Scaffolded Framework for AI Integration in Computing Education (pdf).

Design Moves

• Scaffolds the course in stages: no AI, then AI as a pair programmer, then freer AI use for an ambitious final project
• Builds independent technical judgment first: in early projects, students write, debug, and reason through their code on their own
• Makes critical evaluation part of the work: once AI is introduced, students stay responsible for checking, refining, or rejecting its suggestions
• Ends with an open-ended project where students apply their foundation and judgment to build something ambitious, using AI as a multiplier rather than a substitute
• Provides a clear AI policy rationale, showing students how AI is, and will be, part of their future careers. His policy tells them: “The more knowledge and skill you bring to the table, the more value you get from using them.”

These moves may translate beyond software engineering: build independent judgment first, practice evaluating AI’s suggestions critically, then apply both to open-ended work. What counts as “foundational” will differ by discipline.

In ENVS 429 Environmental Leadership Program, Professor Boulay asks student teams to co-construct and personalize their own GenAI use policy.

Her prompt to students: "What is your team policy on GenAI use? Read the syllabus statement on GenAI, then have an open, empathetic conversation to create a consensus policy for your team's approach to GenAI within the syllabus boundaries."

An example student team statement: "We will refrain from using artificial intelligence (AI) unless it is for the purpose of finding resources, brainstorming ideas, summarizing information or articles, creating outlines, or proofreading our work. AI may not be used in place of our own words or work. If AI is used, we must properly cite it to avoid plagiarism. Any suggested topics will be fact-checked and further researched by the team members."

Design Moves

• Provides her own general GenAI policy for the course.
• Asks student teams to write their own versions of the policy as part of setting their group norms.
• Has students report on how they adhered to, or revised, their norms during the term.

Professor Heffernan narrates her AI-related teaching choices out loud to students, activity by activity, rather than setting one blanket course policy. The same class might run AI-free one day and AI-assisted the next. What stays consistent is that she explains why, tied to what that specific activity is trying to teach.

She says, "I am modeling and teaching discernment about purpose."

Design Moves

• Explicitly narrates why each activity is designed the way it is, given that AI "is sitting right there," making the pedagogical reasoning visible in the moment.
• When the purpose is peer conversation and high engagement, runs hands-on, closed-computer activities built from printed readings; AI is deliberately absent.
• When the purpose is a task like rewriting questions to a different level of Bloom's taxonomy, gives direct instruction on how to use and cite AI; AI is deliberately present.
• Frames the throughline for students as discernment about purpose, not a fixed yes/no rule to memorize.

This practice also fits the Promote Process teaching move. See more of Prof. Heffernan's approach and other examples of promoting process on the Faculty Examples: Promote Process page.

Professor Gopalan's EDLD 652 syllabus states her AI policy in specific, task-level terms rather than one blanket rule, naming exactly where AI is welcome and where it isn't, and committing to keep flagging that distinction as the course goes. Prof. Gopalan says, "Their use will be allowed and encouraged for certain activities and tasks in this course (e.g., as a copilot for coding during lab activities and problem sets, and for the final project), which I'll explicitly note along the way. I'll also clearly indicate when such systems are not allowed or are strongly discouraged."

Design Moves

• States the policy in task-specific terms (coding labs, problem sets, final project) instead of one course-wide rule.
• Commits to flagging AI-allowed vs. AI-discouraged moments "along the way," so the policy stays visible at the point of use, not just on day one.
• Frames recognizing AI's "pluses and minuses" as a stated learning goal of the course, not just a compliance rule.

This practice also fits the Build AI Literacy teaching move. See more of Prof. Gopalan's approach and other examples of effective student communication on the Faculty Examples: Build AI Literacy page.