6Professionalism
10Information Literacy
AI Literacy: Beware of the Creep
Did you cross the line, and did you notice?
This happens to almost everyone, often without them noticing.
A student starts a paper by asking AI for a few topic ideas. That feels reasonable. Brainstorming is exactly what AI is good for. Then they ask for an outline. Still feels fine. Then they're stuck on the introduction, so they ask AI to write a sample one "just to see what it might look like." It sounds pretty good, so parts of it end up in the draft.
Later, worried the writing isn't polished enough, they ask AI to improve a paragraph. Then to make it sound more professional. Then to rewrite the conclusion.
By the time the paper is done, they honestly can't say how much of it is theirs and how much came from AI. They weren't trying to cheat. The line moved one small step at a time, and nobody was keeping track of where it ended up.
This is AI creep: AI's role in your work growing a little at a time, where each step feels small and reasonable, but the end result is very different from where you started.
AI creep: the gradual, often unnoticed growth of AI's role in your work.
The AI Creep Meter
Each stage on this scale means AI plays a bigger role in producing the actual content of your work.
The AI Creep Meter
- Brainstorming
- Outlining
- Drafting
- Revising
- Rewriting
- Submitting
Lower stakesHigher stakes
Brainstorming and outlining are usually lower stakes. You're still doing the thinking, and AI is helping you organize it. As AI moves into drafting, revising, and rewriting, it does more of the work for you, and the line between "my work" and "AI's work" gets harder to see.
That doesn't mean the right side of the meter is always off-limits. That depends on the assignment and your instructor. It means the farther right you go, the more important it is that you can explain what happened, and that you chose to go there on purpose.
Lesson 1.4: each assignment's AI use label tells you how far along the meter you're allowed to go.
Set your boundary before you start
The fix is deciding, before you open the chat, what role AI is allowed to play in this piece of work. Tools have jobs. A hammer is great for nails and useless for screws. AI is the same: useful for some jobs on a given assignment, the wrong tool for others.
Before you start, ask yourself:
- What am I trying to learn from this assignment?
- What parts of this work need to be mine, in my own words and my own thinking?
- What role, if any, should AI play here?
Then tell the AI that role directly. A few examples:
- Tutor
Explain this concept in simple terms, but don't complete the assignment for me.
- Brainstorming partner
Help me generate ideas and questions to explore. Don't write content I can submit.
- Study coach
Ask me questions that help me think this through myself.
- Editor
Give me feedback on clarity and organization, but don't rewrite my work.
Setting the boundary up front does two things: it keeps AI in its lane, and it gives you something to point back to if you're asked how you used it.
Why this connects to BE Honest
AI Literacy BE Honest
BE Honest is less about getting caught than about whether your work is worthy of your name. AI creep makes that question hard to answer, even when you aren't hiding anything, because you've lost track yourself.
If an instructor asked you right now, "How did you use AI on this?", could you answer clearly? If the honest answer is "I'm not totally sure," notice that, because it tells you the boundary slipped.
What attribution actually looks like
"Be transparent about AI use" can feel vague until you see it written out. Attribution doesn't need to be long or formal. It needs to be specific enough that someone else could tell what AI did and didn't do.
"I used AI to brainstorm three possible topics and to help me build an outline. The writing, examples, and conclusions are my own."
"I drafted this myself, then used AI to revise my conclusion paragraph for clarity. I reviewed the changes and adjusted the wording so it still sounds like me."
Both examples name what AI did (brainstorm, outline, revise one paragraph), and both are specific enough that the student could explain more if asked. Neither says "I used AI" and stops there. A vague disclosure is hard to act on, for you or your instructor. A quick way to write your own: name the stage or stages on the Creep Meter where AI was involved, then say what stayed yours.
Keep a trail in Lesson 1.4 makes attribution much easier.
Before you submit, ask yourself
- Can I explain my work and how AI helped me? (That's the third of the 3 Qs.)
- Could I show my process if asked?
- Does this still show my own learning and thinking?
- Would I be comfortable talking through my AI use with my instructor?
If any answer is no, that's the moment to pause, not after you've submitted.
In your own life
- Think of a time you used AI for a task, for school or anything else. Where would that use land on the AI Creep Meter?
- Has your AI use ever gone further than you planned, without you really deciding it should? What did that feel like once you noticed?
- What would it look like to set a boundary with AI before your next assignment, and tell the AI what that boundary is?
The first question here is Question 4 in Mile Marker #4.
AI policies are still evolving, and they'll keep changing. What won't change: transparency, honesty, and ownership of your learning are yours to decide, every time.
The most important decision about AI use is the boundary you set at the beginning of a project, not anything you do at the end.