The Honest AI-Use Spectrum
"Is using AI on this cheating?": let's stop pretending it's a yes/no question.
The conversation about AI and writing is broken. Half the people you talk to say "AI is just a tool, use it for everything." The other half say "AI is cheating, never touch it." Both are wrong.
The truth is that "using AI on a paper" isn't one thing. It's a spectrum of at least seven different things that all get crammed under the same scary label. Some of them every professor on earth is fine with. Some of them will get you suspended. Most are in the middle, and the middle is where you need a real framework.
This module is going to give you that framework. Today's lesson is the map. The next nine lessons are the moves you make on it.
How to use the spectrum
Once you can locate where on this map a given task lives, you'll know what's allowed, what's risky, and what to ask your professor about.
The seven steps of writing: and where AI sits at each one.
Every paper you write moves through (most of) these stages. AI can plug in at any of them. The honesty stakes change a lot between them.
1 · Understanding the assignment · GREEN
- What it looks like: "Here's the prompt. What's the professor actually asking? What kind of paper is this: argumentative, analytical, reflective?"
- Universally fine. No professor cares that you used AI to make sure you understood the assignment.
2 · Brainstorming angles · GREEN
- What it looks like: "Here's the prompt. Generate eight angles I could take. Flag the clichés."
- Almost universally fine. Brainstorming is the part professors are most relaxed about, partly because it's identical to what you'd do at office hours, on a walk with a friend, or in a study group.
3 · Outlining the argument · LIGHT YELLOW
- What it looks like: "Here's my thesis. Help me structure intro / four body sections / conclusion."
- Mostly fine, with a caveat. Most professors are OK with this if the thesis is yours. If you handed AI a vague topic and let it pick your thesis, you've drifted into something more substantive. Your argument is the thing your paper is supposed to be.
4 · Writing the actual draft · YELLOW (the contested middle)
- What it looks like: "Write me three paragraphs about X," and then you paste them in.
- This is where most policies draw the line. Even most "AI-friendly" classes don't allow you to submit AI-written paragraphs as your own work without disclosure. Lesson 4.5 is about how to get help with drafting that's actually allowed.
5 · Editing your own draft · GREEN (the workflow professors love)
- What it looks like: "Here's my draft. Don't rewrite. Tell me where my argument is weakest, where I'm being repetitive, where the evidence doesn't support the claim."
- A widely-allowed move. Lesson 4.5 is built around this. AI is a better editor than it is a writer, and editing your own draft is a skill that makes you better.
6 · Polishing and proofreading · GREEN
- What it looks like: "Catch typos, fix grammar, flag awkward sentences." (Plus the AI-pattern scrubbing prompt from Lesson 4.6.)
- Universally fine. Spellcheck has been allowed for 30 years; this is spellcheck on steroids.
7 · Submitting AI-generated work as your own · RED
- What it looks like: Pasting an AI-written paper, adding your name, hitting submit.
- This is plagiarism. Not because of the AI, but because you submitted work that wasn't yours. The same rule that's always applied to copying anyone else's work.
The bright line you can defend in a hallway conversation.
You'll hear different versions of the AI rule from different people. Here's a single test:
The Hallway Test (the bright line for this whole module)
Could you explain every paragraph of your submitted writing to your professor in a five-minute hallway conversation?
Could you say where the argument came from? Why this evidence? Why these words? If yes, you're on the right side of the line. If no, you've crossed something. The professor is asking if the thinking is yours.
This test maps cleanly onto the spectrum: brainstorming, outlining, editing, polishing all leave the thinking with you. Letting AI write the paragraphs gives the thinking away.
Plagiarism rules existed before AI, and the underlying question (can you stand behind this work as your own thinking?) was always the real question. AI just made the easy way to cheat much easier. The honest path didn't change.
How to use this map for the rest of the module.
Here's how the rest of Module 4 plugs into the spectrum:
The map of the module
- 4.2: Find out what your specific school and your specific professor say about all of this. (Their rules beat any spectrum.)
- 4.3: How to use AI well at the green-zone steps (brainstorming + outlining).
- 4.4: How to teach AI your actual voice so anything you generate sounds like you, not like default.
- 4.5: The Edit-Don't-Write Workflow, the centerpiece of the whole module. Yellow-zone drafting decisions, the green-zone editing pass, and the process trail you'd show a professor. Three lessons folded into one workflow.
- 4.6: How to spot AI-style patterns in your own draft.
Honest Work Code · Rule 3: "Their rules beat my rules."
The spectrum on this page is the general case. Your professor's rules (and your school's rules) beat this map every time. If your prof says "no AI at all on this paper," that's the rule, even if my map says outlining would normally be fine. If your school's policy is more permissive, still follow the prof. The next lesson (4.2) is about how to find out what those rules actually are.
Up next: 4.2: Find Your School's Actual Policy
Half the AI-policy panic in your life is because you don't actually know what your school says. Most students have never read it. We're going to fix that, and give you a professor-email script for when the policy is unclear.
Continue to 4.2 — Find Your School's Policy →