When students use AI writing tools, they face a new challenge that extends beyond traditional writing instruction. The question is not simply whether AI helps them write better, but whether they can maintain authorial control while using these tools. The VAPOR framework offers a practical approach to this challenge, one that applies across disciplines and refocuses student attention on the rhetorical decisions that define authentic composition.
The Problem with Current Approaches
Most institutional responses to AI in writing fall into two categories: prohibition or passive acceptance. Neither approach addresses what students actually need—a systematic way to evaluate their relationship with AI as they write. Students report feeling uncertain about which AI suggestions to accept, worried that their writing no longer sounds like them, and unclear about where their ideas end and AI assistance begins. These concerns cut across business memos, lab reports, and personal essays alike.
The VAPOR framework (Voice, Agency, Posture, Origin, Responsibility) emerged from recognizing that AI literacy in writing is fundamentally about rhetorical awareness.
Students who understand how AI makes persuasive choices are better equipped to use it critically, regardless of their major or writing task.
What VAPOR Measures
The framework operates through five interconnected dimensions:
Voice asks whether the writing sounds authentically like the student. This matters in a biology lab report just as much as in a reflective essay. When students can identify which sentences feel “off” or unlike their typical writing style, they develop awareness of how AI shifts their rhetorical presence.
Agency examines whether students drove the AI interaction or passively accepted its output. Did they make deliberate choices about which suggestions to accept based on their audience and purpose? Or did they click “accept all” without rhetorical reasoning?
Posture addresses strategic positioning. Students consider whether their writing reflects their intended relationship with the topic and audience, or whether they are simply performing “expected” academic language. This dimension helps students distinguish between strategic formality and hollow performance.
Origin focuses on traceability. Can students identify which parts of their writing came from AI? Do they remember their original ideas before AI interaction? This dimension addresses a practical concern: students often lose track of their thinking when AI generates substantial text.
Responsibility measures students’ confidence in defending their work. Would they feel prepared to explain their claims if challenged? Do they understand the concepts well enough to rewrite them in multiple ways? This dimension connects directly to academic integrity without relying on punitive framing.
Implementation Across Contexts
The framework functions as a self-assessment tool that students complete at strategic points in their writing process. A student working on a nursing care plan might use VAPOR before submitting to verify they can explain their clinical reasoning. An engineering student drafting a project proposal might check whether AI suggestions align with technical specifications and professional standards in their field.
Higher VAPOR scores indicate greater risk; showing low confidence in student voice, agency, or responsibility. Lower scores suggest critical engagement with AI assistance. The scoring itself matters less than the reflection it generates. When a marketing student realizes they cannot explain the market analysis AI helped them write, or when a psychology student notices their literature review lacks their interpretive voice, the framework has achieved its purpose.
Faculty across disciplines can adapt VAPOR questions to fit specific genres and assignments. The core questions remain the same, but the reflection prompts can reference discipline-specific concerns. A computer science instructor might ask students to trace their algorithmic thinking before and after AI assistance. A history professor might focus on whether students can defend their interpretation of primary sources regardless of how AI helped them structure arguments.
Why This Matters Beyond Composition
AI writing tools impact every field where students must communicate their knowledge. The VAPOR framework recognizes this reality by focusing on transferable critical thinking rather than genre-specific rules. Students learn to ask: Does this represent my thinking? Can I explain how I got here? Will I stand by this work?
These questions apply whether students are writing research proposals, technical documentation, patient notes, or policy briefs. The framework does not prescribe when AI use is appropriate. That question remains a disciplinary and contextual decision. Instead, it gives students and faculty a shared language for discussing the rhetorical implications of AI assistance.
Recent research confirms that AI-generated text often lacks the authorial markers that characterize human writing, particularly the strategic positioning and voice development that signal genuine understanding. The VAPOR framework addresses this gap by making these markers visible to students and instructors. When students actively work to maintain their voice and agency, they develop stronger critical AI literacy across writing situations.
Practical Integration
Instructors can integrate VAPOR at multiple touchpoints: before students begin using AI (to establish awareness), during drafting (to check alignment with rhetorical goals), and before submission (as required reflection). Some faculty may use it for peer review, asking students to predict each other’s scores and discuss discrepancies. Others might incorporate it into portfolio reflection, tracking how student scores change across the semester as they develop more sophisticated AI collaboration strategies.
The VAPOR Self-Assessment for AI-Assisted Writing model (below) includes adaptation recommendations for discipline-specific modifications and a reflective scoring sequence for students. Faculty decide when students complete the assessment, how it integrates into course activities, and what level of reflection to require. The framework works across disciplines because it focuses on transferable rhetorical principles rather than genre-specific writing conventions. While its origin stems from writing applications, the framework can extend to other forms of AI-assisted student work like presentations, data analysis, problem-solving, and coding where authorial control and critical thinking matter.
The framework also serves programmatic assessment needs. Departments can use aggregated VAPOR data to identify where students struggle most when co-composing with AI. Do students struggle to maintain voice, understand origin, or take responsibility for AI-assisted content? These patterns inform curricular decisions about where to strengthen instruction in critical AI use.
Looking Forward
As AI writing tools become more sophisticated, the need for critical frameworks grows more urgent. Students will encounter these tools throughout their academic and professional lives. The VAPOR framework equips them to use AI as a rhetorical tool while maintaining the authorial control that defines genuine composition.
This approach shifts the conversation from “catching” AI use to developing students’ capacity for critical judgment about when and how to integrate AI assistance. That shift matters for every discipline where writing serves as evidence of thinking, which is to say, nearly all of them.
References
Beck, R. J., Skinner, E., Lopez, A., Montgomery, K. L., & Yancey, K. B. (2024). Generative AI in first-year writing: An early analysis of affordances, limitations, and a framework for the future. Computers and Composition, 71, 102827. https://doi.org/10.1016/j.compcom.2024.102827
Carter, J., & Dousay, T. A. (2024). Writing with generative AI and human-machine teaming: Insights and recommendations from faculty and students. Computers and Composition, 71, 102833. https://doi.org/10.1016/j.compcom.2024.102833
Plate, D. (2025). Writing as curation: Empowering authorial agency in AI-assisted composition through style prompting and quotation glosses. International Journal of Emerging and Disruptive Innovation in Education: VISIONARIUM, 3(1). https://doi.org/10.62608/2831-3550.1033
Singh, J. K., Daniel, B., & Koh, J. (2025). Empowering authorship with AI: A novel academic writing technology for authorial voice. International Journal of Artificial Intelligence in Education, 35, 3406-3428. https://doi.org/10.1007/s40593-025-00503-8
Useche, A. (2025, October). Using Bloom’s taxonomy to understand AI adoption in higher education. OLC Insights. https://onlinelearningconsortium.org/olc-insights/2025/10/blooms-for-ai-adoption/
Zhang, Y., & Ma, Y. (2025). The impact of generative AI on academic reading and writing: A synthesis of recent evidence (2023-2025). Frontiers in Education, 10. https://doi.org/10.3389/feduc.2025.1711718
Dr. Sarah Martin is Associate Teaching Professor and Faculty Head in the College of Integrative Sciences and Arts at Arizona State University. She works with faculty to design online learning experiences that support effective writing and communication instruction in professional and technical contexts. Dr. Martin has previously led enterprise-level communication projects for the federal government and directed award-winning professional communication programs at R1 institutions. She holds an M.B.A. from the Naval Postgraduate School and a Ph.D. in Technical Communication & Rhetoric from Texas Tech University.
