Rethinking Writing Support: AI as Peer, Editor, and Partner in the Undergraduate Classroom

• In Event: Generative AI and Academic Writing: Studies from Diverse Contexts

Abstract

The emergence of AI in the public sphere has undoubtedly changed the future of education. For many teachers and writing instructors, this change is top of mind, often guided by fears of academic dishonesty and the devaluing of writing skills and instruction (Cotton, Cotton, & Shipway, 2023; Fütterer et.al., 2023). However, the actual impact AI will have on writing instruction and students’ experience in the writing classroom is still largely unknown. Early studies explored theoretical applications or gathered student and teacher responses, but their designs were often speculative rather than grounded in real-world classroom integration. (Liang, Stephens, & Brown, 2025; Marín, Bond, & Gouverneur, 2019). Our study addresses this gap by examining how AI tools are experienced in practice and how they shape the writing process, particularly when integrated into course design rather than used as an external supplement. By employing case study methodologies, we analyzed student coursework, AI-human interaction transcripts, interviews, and classroom observations. We identified distinct patterns in students’ methods of collaboration with AI, as well as reflections that reveal how students approach AI use and how it may shift the writing process for postsecondary learners. Our findings highlight the different collaborative roles students assign to AI systems, often mirroring those of human collaborators, while also demonstrating how students’ values, ethics, and sense of autonomy persist throughout their writing process, even when AI tools are fully integrated.
We begin by analyzing the method of integration and the resulting collaborative roles that shaped students’ interactions with AI. Inspired by the potential for AI to serve as a writing partner rather than a replacement for student effort, the course design offered structured guidelines for engagement while still allowing students the autonomy to shape their interactions with AI and define the role it would play within their writing partnership. From our analysis of these interactions, we found that students rotated among three distinct collaborative roles for AI: Teacher/Tutor, Peer Editor, and Thinking Partner. We will explore the defining characteristics of these roles and their implications for the future of collaborative writing instruction.
We then explore the themes of time and autonomy, which surfaced as meaningful patterns, and offer insight into how students’ sense of value and ethics shape their approach to writing. Students frequently noted that the tools were beneficial time savers, but pushed back against these time-saving methods in areas where they prioritized their own voice or ideas. Similarly, many students expressed a strong sense of ethical boundaries with the tools, even in the absence of explicit guidelines, and often qualified the perceived usefulness of AI by their ability to maintain individuality and control over the output. These findings challenge dominant concerns that frame AI as a threat to student learning and academic integrity. Instead, our study suggests that when thoughtfully integrated, AI can foster new forms of collaboration while reinforcing students’ commitment to authorship, voice, and ethical responsibility. We will conclude with a discussion of the pedagogical implications of these findings for writing educators navigating the age of AI.

Author

• Maura White, University of California - San Diego