Whose Words Are These? University Faculty Perspectives on Originality and AI-Mediated Composition
Education Science (Under Review), 2026
What This Paper Is About
Generative AI has blurred the line between “student-written” and “AI-generated.” This paper captures how higher education instructors are renegotiating what counts as authentic student work — not through policy documents, but through their lived classroom experiences.
Core Argument
The paper argues that authenticity in student writing is not binary (human vs. AI) but exists on a spectrum. Instructors are developing nuanced, often implicit frameworks for evaluating student work that go beyond simple plagiarism detection:
- Process over product — some instructors now value the writing process (drafts, revision history) more than the final output
- Voice recognition — experienced instructors can detect mismatches between a student’s in-class voice and their submitted writing
- Meaningful engagement — the question shifts from “Did you write this?” to “Did you learn from this?”
What Instructors Actually Do
| Strategy | How It Works |
|---|---|
| Scaffolded assignments | Breaking writing into observable stages to ensure genuine engagement |
| In-class writing | Low-stakes writing that establishes each student’s baseline voice |
| Reflective components | Requiring students to explain their thinking and revision choices |
| AI-transparent policies | Allowing AI use but requiring disclosure of how and where it was used |
Why It Matters
Most research on AI and academic integrity focuses on detection (can we catch AI use?) or policy (should we ban it?). This paper focuses on something more fundamental: how instructors are redefining what “learning” looks like when AI can produce competent text on demand. Their strategies reveal an emerging consensus that the goal of writing assignments is not the text itself, but the cognitive work behind it.