Adobe Stock Creative Commons Image Not AI Generated
Different teachers take different approaches to whether—and how—generative AI should be used in the writing classroom. SWLF instructor J.T. Bushnell, however, makes a compelling case in the Winter 2025 issue of Writer’s Chronicle for meeting students “where they’re at.” Increasingly, J.T. argues, that place is an AI prompt window, waiting for an answer.
J.T. cites eye‑opening statistics on student AI use, including a Chronicle of Higher Education survey showing that 88 percent of undergraduates are using AI for coursework. “Telling people to just say no isn’t working,” he writes. “We have to find better ways.”
Alongside fellow SWLF instructor Wayne Harrison, J.T. explored what some of those “better ways” might look like through a study on AI use in the creative writing classroom. Funded by the Ecampus Research Unit, the study involved 31 introductory creative writing students at OSU, who completed writing tasks under three conditions: without AI; with ChatGPT but without guidance; and with ChatGPT paired with explicit instructor guidance.
The results suggest that AI can increase creativity only when instructors provide clear direction. Unguided AI use tended to flatten originality—helping weaker writers but diminishing stronger ones—and reduced overall student satisfaction. As Wayne explained in a press release, “This approach has huge implications, not just for writing classes, but for any discipline where critical or creative thinking matters. We don’t have to choose between banning AI entirely and throwing up our hands in defeat.”
Instead, the study points to a more nuanced approach that emphasizes focused, intentional use of AI during idea generation and revision, rather than during the core act of composition. J.T. offers a concrete example from fiction writing: the common trope of beginning a story with a character waking up. “If you ask AI, ‘What’s one aspect of this excerpt that is bland, cliché, or unoriginal?’” he notes, “it’ll zero in on places like this.”
Similarly, a student might ask AI to identify where a description of a setting lacks clarity or needs more detail, or to suggest possible professions for a character when feeling stuck. What’s critical, J.T. and Wayne argue, is that students come to understand how human discernment—choosing from, reshaping, or rejecting AI suggestions—shapes the creative process itself. J.T. refers to the heart of that process as “the trance, the turquoise pool,” adding, “If you’re like me, you won’t let AI anywhere near it.”
It’s easy to assume that allowing AI into the creative classroom—even in limited ways—will unleash a flood of AI‑generated writing, the temptation proving too strong for students to resist. In practice, Wayne and J.T. have found the opposite. In an honors course built around using AI in fiction writing, Wayne notes that students “learned to use it as a practical tool,” while remaining “deeply protective of their own creative instincts, and suspicious of letting AI cross into the part of writing they see as fully theirs.” J.T. reports a similar result: the pedagogies he developed with Wayne help students recognize how often AI‑assisted techniques fall short. “Rather than fold these techniques into their process,” he says, “students largely turn against them—which is a better outcome than I ever hoped for.”
The study is one contribution to a much larger and ongoing conversation about teaching, writing, and generative AI. More research is needed, as is greater institutional support for educators navigating this shift. In his role as a Critical AI Literacy Fellow at the Valley Library during the summer of 2025, J.T. helped address that need by developing a suite of freely available resources—including instructional videos and a learning module—designed to support instructors seeking thoughtful, creative, and critical ways to engage AI in the classroom.