Module 313 min read · AI for Students

Writing with AI: Where to Use It and Where Not To

Writing is the activity most dramatically changed by AI for students — and the one where the risks of misuse are most serious. Every major university is wrestling with how to handle AI in written assignments. This module gives you a clear, principled framework for using AI in your writing process in ways that genuinely serve your learning and maintain your integrity.

What happens when AI writes for you

When AI writes your essay, a specific transfer takes place: the thinking process moves from your head to the model. That process — the struggle to articulate an argument, the discovery that your initial thesis doesn't hold, the satisfaction of finding the precise word — is the actual educational content of a writing assignment. The document is evidence of that process, not the point of it.

So when students use AI to write entire papers, they're not just risking academic integrity consequences. They're forfeiting the development of skills they actually need — skills that are increasingly valuable precisely because AI can handle the mediocre work, raising the bar for what genuine human reasoning is worth.

The policy minefield

AI policies vary enormously by institution, department, and instructor. Some explicitly prohibit AI use in any form. Others allow it with citation. Many are somewhere in the middle or still figuring it out. There is no universal rule — you must know the specific policy for each assignment. "I didn't know AI wasn't allowed" is not a defense that works in academic integrity proceedings.

The writing uses that are almost always acceptable

Brainstorming and idea generation
Asking AI to generate 10 angles on a topic, list potential counterarguments, or suggest what kinds of evidence would support a claim — this is like talking to a very well-read brainstorming partner. You're not producing work; you're generating material to react to.
Structural feedback on your draft
Share your own draft and ask: "Is my argument logically coherent?" or "Does my evidence actually support my thesis?" or "Where does this lose the reader?" This is editing assistance — valuable and fundamentally different from having AI write the content.
Grammar and clarity checking
Running your completed draft through AI for proofreading is analogous to using spell-check — it's a tool for improving your own writing, not replacing it. Most policies allow this. Verify yours does.
Explaining writing concepts
If you don't understand what "transition sentences" are, or how a literature review differs from an annotated bibliography, asking AI to explain is legitimate. Understanding what you're supposed to do is different from having AI do it for you.

The writing uses that are problematic

  • Asking AI to write sections you'll paste into your paper — even if you edit them afterward, the fundamental authorship is compromised in most policy frameworks
  • Asking AI to "improve" your draft in ways that produce substantially different text — at some point, the voice and content become AI's, not yours
  • Having AI generate your thesis — the thesis is the center of your argument; if AI generates it, what exactly did you contribute?
  • Using AI-generated content without disclosure when disclosure is required — academic integrity violations are permanent record events in most institutions

Developing your own writing voice

There is a long-term cost to relying heavily on AI for writing that goes beyond academic risk: you don't develop your own voice. A written voice — a distinctive, recognizable style that reflects how you think — takes years to build and requires actually writing, a lot, without assistance. Students who outsource their writing in college arrive at careers unable to communicate in their own voice, dependent on tools that may or may not be available in professional contexts.

A practical writing workflow with AI

1. Write your first draft entirely yourself — no AI. It will be rough. That's fine.

2. Ask AI for structural feedback on what you wrote. Take notes on what it identifies.

3. Revise your draft based on the feedback — you doing the rewriting, not AI.

4. Ask AI to check for clarity issues in specific sections.

5. Do a final proofread.

This process is faster than writing from scratch, keeps your voice intact, builds your skills, and stays clearly within most acceptable-use policies.

When AI can improve your writing skills

Used thoughtfully, AI can actually accelerate your development as a writer. The key is using it in a reflective, deliberate way:

  • After AI identifies a structural weakness in your paper, understand why it's a weakness — don't just fix that paper, learn the principle
  • If AI suggests clearer phrasing for a sentence, study the difference — what makes the new version clearer? Apply that principle to your next draft
  • Ask AI to explain what makes a paragraph you wrote particularly strong or weak — this gives you explicit feedback on your own patterns
The long-term advantage

Students who use AI as a learning tool rather than a replacement tool will be better writers at graduation than students who didn't use AI at all — because they got higher-quality, more frequent feedback on their own writing. The goal is to make AI serve your development, not to substitute for it.