Field Guide Entry — A student decision framework for AI use

Talking to Students About It

A student decision framework for AI use

This provides students with the tools to decide whether AI assistance is supporting their learning or actually causing them to bypass learning altogether.

The concept

While making an AI policy for your students will set clear guidelines for how AI should and should not be used in your class, it doesn’t help students with the inevitable decision of whether to use AI in the future. If you want to help your students navigate the AI landscape beyond the confines of your class, it might be useful to provide them with a decision framework as well. Not only does this help build a healthy view of what AI should be used for, it builds the habit within students to ask why they’re actually completing tasks and work for the course.

This entry is for faculty looking to create a decision framework for your students. The AI prompt below generates a discipline-specific draft. The design guidance helps you refine it into something that reflects what your course is actually trying to develop.

AI prompt — build the framework

Step 1 — generate a discipline-specific draft

I teach [course name and level]. I want to give students a decision framework — a short set of questions they can ask themselves before using AI on any assignment — that helps them distinguish between AI use that supports their development and AI use that substitutes for it. The framework should be specific to what [your discipline] actually values: the skills it builds, the kinds of thinking it develops, and what “doing this well” requires. It should not be a list of rules about what’s allowed or banned. Draft 4–6 questions that a student could genuinely use as a self-check. For each question, briefly explain what it’s testing for and why it matters in this field specifically.

Step 2 — stress-test it

Now play the role of a student who wants to use AI for every part of this assignment and is trying to rationalize doing so. Work through the framework question by question and show me where a motivated student could answer “yes” to everything while still doing work the framework was designed to flag. Then tell me which questions are most resistant to rationalization and which need to be tightened.

A base framework — five questions for any course

Below is a generic example of what a decision framework might look like. These questions are a starting point. Use the AI prompt above to adapt them to what your specific course is developing.

What skill is this assignment developing in me?

Can name it specifically

Clearer signal: AI assistance that doesn’t touch this skill may be fine.

Can’t answer this

Stop here. If you don’t know what you’re developing, you can’t evaluate whether AI undermines it.

Have I made a genuine attempt on my own first?

Yes — I have a draft, notes, or attempt

Using AI to react to your own work is usually fine. You’re steering it; it’s not replacing you.

No — I’m starting with AI

This is the highest-risk scenario. Starting from AI output means you may never encounter the difficulty the assignment was designed to develop.

If AI does this part, what am I practicing instead?

Something the course also values

Acceptable tradeoff — as long as what you’re doing is genuinely substantive, not just reviewing output.

Nothing — I’m reviewing, not doing

Reviewing AI output feels like work. It often isn’t. Be honest about whether you’re producing or just approving.

Am I using AI to extend my thinking, or to start it?

Extending — AI is responding to my ideas

Your thinking is upstream. AI is downstream. This is the direction that tends to build skill.

Starting — AI is generating the ideas

This is where the cognitive work that builds expertise is most likely to be skipped.

Could I explain my process and reasoning to my professor?

Yes — I could walk through every decision

Good signal. Intellectual ownership usually means you could explain it.

No — I couldn’t account for some of it

This is the clearest sign something went wrong. If you can’t explain it, you don’t own it.

Judgment zones — what the framework is pointing toward

Generally appropriate

AI as a coworker

Critiquing a draft you wrote. Explaining a concept you’re stuck on. Checking your reasoning after you’ve reasoned. Generating examples after you’ve named the principle.

Depends on the course

AI as a scaffold

Outlining before drafting. Generating options before selecting. Summarizing sources before synthesizing. Whether this is fine depends on whether the scaffolded step is what the course is developing.

Typically problematic

AI as the author

Starting from AI output. Submitting AI-generated work with light editing. Using AI to perform the exact reasoning, writing, or problem-solving the assignment is designed to develop.

Building your version — how to use this entry

Start with what your course is actually developing

Before writing a question, name the skill. Try to be as specific as possible. How is that skill being applied? What is the context? The more specific you are here, the better the framework will be.

Useful starting questions to ask yourself: “What can a student who passes this course do that they couldn’t before?” / “What does doing this discipline well actually require — what’s the hard part that can’t be shortcut?” / “If a student used AI for everything and still got an A, what would be missing?”

Use the AI prompt to generate a discipline-specific draft

Keep in mind that the AI output will often give you a nice starting point, but it is up to you to refine the framework. It is okay to disagree with what the AI is suggesting. This draft is just a tool to help up clarify our thoughts.

Trim to 3–5 questions and make it student-readable

Avoid putting too much into the framework. Students won’t use it if there are too many steps. These should feel like natural questions of curiousity and not a strict rubric.

Introduce it with the reasoning, not just the list

When you share the framework with students, explain what it’s for. Some students will appreciate the exercise more if they know why they’re doing and why you’re taking time out of class for it.

Caveats

A decision framework doesn’t detect or prevent misuse. A student determined to work around it can. But students who are looking for real AI guidance will benefit from this decision-making exercise.
Students may answer “yes” to all five questions and still produce AI-dependent work. The stress-test prompt above is specifically designed to find where this happens.
The framework should be consistent with your actual policy. If you say “don’t use AI” but also offer a framework for when AI use is appropriate, students will notice the contradiction.

Reflection questions

What question in the base framework is least applicable to your course — and what does that tell you about what your course is developing?
Where could a motivated student in your course answer “yes” to every question while still doing work that undermines the course’s goals?
Have you introduced the framework with reasoning, or just distributed it as a list? What difference would that make?
Is there an assignment in your course that you’d struggle to explain to a student using this framework? What does that reveal?

5.2-A student decision framework for AI use