Study: College Students Edit AI Text Rather Than Simply Using It
Undergraduate writers are not simply letting artificial intelligence write their assignments. A pilot study of 20 students at Kennesaw State University found that most students actively edit and reject AI-generated text, treating the tool more as a brainstorming partner than a shortcut.
Researchers used think-aloud protocols - a method where participants verbalize their thoughts while working - to observe how students actually compose when AI tools are available. They recorded screen activity and voice transcripts during 20-minute writing sessions, then analyzed the data without being present to avoid changing student behavior.
How students actually use AI while writing
Students turned to AI most often at the beginning of the writing process to generate ideas or draft a thesis statement. One student described the approach this way: "After [generating a few ideas,] I usually just use that [output] as a prompt." The AI output served as a starting point, not a final answer.
After generating initial ideas, most students continued drafting independently. Transcripts included statements like "I think my thesis should be …" or "Let me write this part," showing students retained control over their arguments.
When AI text appeared in drafts, students rarely accepted it unchanged. They actively revised the language, rewrote passages, and evaluated whether the output matched the assignment requirements. One student redirected the AI when it strayed from the prompt: "AI is not following the prompt … try again."
Some students rejected AI suggestions entirely. When AI responses felt too generic or disconnected from their argument, students switched back to their own writing. Several explicitly decided not to use AI for research or other parts of the assignment.
Students also turned to AI during moments of uncertainty. As one participant said, "I used a lot of AI because I was struggling." Even then, students used it as support while drafting rather than copying and pasting directly.
What educators should know
The findings suggest that generative AI enters student writing as a negotiated collaboration, not as a replacement for human authorship. AI most often appears during idea generation, revision, and moments of writer's block, while students maintain control over argument choice, voice, and final phrasing.
Understanding how students actually decide to use AI during the writing process - rather than just examining finished papers - may help educators design assignments and policies that keep the writer in charge.
The research team is expanding the study to 100 undergraduate participants to test whether these patterns hold at a larger scale. The expanded work will also examine how neurodivergent writers interact with generative AI during composing, an area that remains largely unexplored.
For writing professionals, understanding how your peers and colleagues approach AI tools during composition can inform your own practices. Learn more about AI for Writers and how to integrate these tools effectively into your workflow.
Your membership also unlocks:
