Maggie: Can I be crazy about the GAI stuff lol
Jen: Of course
Maggie: Do you remember 3 years ago when people were up in arms about CourseHero???
Jen: Yes
Maggie: Partly bc of the AI-training implications of a huge repository of student and teacher writing ??
Where did that energy go 😭
This text message exchange, from June 4, 2024, marked an early moment that would inform our work on generative AI (GenAI) refusal as it highlighted a concerning disconnect between the field of computers and writing’s critical response to prior predatory technologies like Course Hero, Chegg, TurnItIn, and paper mills and the early response to equally predatory GenAI products like ChatGPT. So, this contribution to the Sweetland DRC Blog Carnival 25 is a polemic, or perhaps more aspirationally, a call-in that challenges the computers and writing community to remember and revisit our past responses to incursions of for-profit technologies in the teaching of postsecondary writing; reject narratives that treat technological progressions as natural and inevitable; and recommit ourselves to more critical approaches that foreground social justice (including linguistic justice) and concerns about power, extraction, and corporate capture—what we refer in this blog post as refusal-in-the-loop writing.
For those who may not recall what Maggie was referring to in the text exchange above, in February 2022, computers and writing scholars—including Jim Ridolfo, Bill Hart-Davidson, Jen Sano-Franchini, and many others—engaged in critique of Course Hero, a course-materials sharing platform that incentivizes uploads through the promise of access. In doing so, they were joining a larger conversation taking place on Medium and academic Twitter about Course Hero’s data practices after it hired digital pedagogy scholar Sean Michael Morris as vice president of academics. Some responded to Morris’s hiring by noting how teachers can work with companies like Course Hero to ensure that our voices inform technological development (sound familiar?). Others, however, took issue with a former edtech critic working with a platform that has been criticized for promoting cheating and for its predatory business model, as Course Hero preys on students, exploits their writing anxieties, and abuses the intellectual property and privacy of teachers.
Because students must upload content to access materials on Course Hero, many teachers voiced concerns that their work had been uploaded without their consent. In Inside Higher Ed, Sano-Franchini explained, “the site uses deceptive design to encourage folks to upload content, but then makes it pretty much impossible to take down. That’s a problem, especially when we consider that Course Hero in a lot of ways targets students who are in need of [academic]support.” Although folks were able to get their content removed from public view after some effort and time, they were unable to have their data deleted from the company’s databases. (It’s worth noting that Course Hero acquired the AI-powered writing product Quillbot in August 2021.)
Nine months later, on November 30, 2022, ChatGPT was released to the public. Given the recent and robust conversation about predatory tech and concerns about academic integrity, intellectual property, and deceptive design, it surprised us that much of the early disciplinary response to GenAI framed these products as the latest in a long line of emergent writing technologies, no different than the word processor, social media, and Wikipedia—especially given our sense of the similarities between the business models of predatory sites like Course Hero and GenAI chatbots like ChatGPT. What’s more, the broad position seemed to be that we should embrace these products as a new writing technology that should be immediately taken up in the college writing classroom (Vee et al., 2023).
What was also unsettling was seeing folks cite Bill Hart-Davidson—usually without deeply engaging with the content of his scholarship—to advance GenAI adoption, given how Hart-Davidson was an important leader in the aforementioned conversations about the incursion of predatory tech in education (Ridolfo, Hart-Davidson, and Lindgren, 2022; Smalley, 2022). Hart-Davidson pushed back hard against companies like Course Hero and Chegg as well as the illicit paper mill industry because of their exploitation of students and students’ writing anxieties. He noted how the requirement that students upload documents for access to content was likely data mining for AI. Given the ways Hart-Davidson’s work took seriously concerns about the relationships among technology, power, politics, and exploitation, and especially the exploitation of students, it’s been disheartening to see folks cite Hart-Davidson’s scholarship on writing with robots as evidence that we should immediately and uncritically adopt and integrate the current wave of mostly corporate GenAI products in higher education.
Much of the field’s early response to GenAI technologies understandably focused on quelling widespread public discourse decrying the fate of students who “don’t read,” “can’t write” or will inevitably fall prey to the irresistible lures of cheating. In response, many writing studies scholars sought to reduce panic with what in many ways amounted to reminders that “we’ve been here before” via analyses of literacy crises and the persistent role of various technologies within writing processes (Byrd, 2023; Graham, 2023; Jamieson, 2022; Johnson, 2023; MLA-CCCC Joint Task Force, 2023; Stanton, 2023). In doing so, these responses frequently contextualized the broad public response to GenAI as part of a historical tendency toward public moral crises about new writing technologies.
Such criticisms of the crisis in response to GenAI have helpfully foregrounded care for students by discouraging the policing of GenAI use in the writing classroom. We also appreciate many of these works for how they raised important issues early in the conversation, including those related to bias and the homogenization of text (Byrd, 2023). At the same time, these responses risk leaning too hard into computers and writing’s techno-optimist tendencies (Gerard, 2006), overlooking the wealth of critical digital scholarship that has importantly grounded the field and that, we believe, better reflects the kind of orientation we should be taking as computers and writing teacher-scholars (Miller, 1979; Hawisher & Selfe 1991; Janangelo, 1991; Selfe, 1999; Banks 2006; Haas 2007; Haas, 2012; Vie, 2013; Sparby 2017; Haas, 2018; Trice & Potts, 2018; DeVoss 2019; Edwards, 2020).
The problems of dismissing concerns about current GenAI products by comparing them to past writing technologies that have been widely adopted in the writing classroom are multiple:
- They frequently advance problematic assumptions about the purported inevitability of widespread (and sustained) adoption of GenAI that have been cooked up and pushed by Big Tech with the backing of trillions of dollars in venture capital. In addition, inevitability discourses are used to forward unsubstantiated claims that we need to start using these products, without concern for how they rely on the extraction of intellectual, material, and land- and water-based resources, via intellectual property theft; global labor exploitation; mineral, energy, and water extraction; and industrial pollution.
- They frequently lead to the dismissal of legitimate, human concerns about the material harms that result from these technologies including issues related to cognitive offloading. Instead, we can take people’s feelings seriously and interpret affectations that may initially read as reactionary, irrational, and “fear-based” repetitions of past resistances to change as examples of our legitimate and highly rational embodied responses to the imposition of increased labor demands, surveillance, and limitations on our agency and ability to opt-out.
- They frequently collapse the histories of several past technologies that are dissimilar in important ways. There are many false comparisons between GenAI and other past technologies that are leveraged to bolster arguments in support of GenAI adoption; GenAI has been compared to calculators, the printing press, the internet, and smartphones. Yet in just about every case, such past communication technologies were (1) created to solve actual communication problems, unlike current GenAI models which were developed in large part by Silicon Valley to further consolidate power and wealth, and (2) saw widespread use because they were reliably useful for meeting specific user needs, not because these products were being forced onto people through coercive design, deceptive marketing, and undisclosed integrations. This is a crucial distinction between current GenAI products and technologies of the past—GenAI constitutes a set of products in search of problems to solve and market share to develop.
With that being said, we call for more critically interrogating the rhetorical uses of history in conversations about GenAI. Put differently, we should be able to recognize the problems of believing that GenAI is more like calculators and word processors—that reliably help users accomplish specific tasks—than predatory technologies that seek to capitalize on entry into the educational industrial complex, like TurnItIn, Course Hero, paper mills, and GenAI. All of these products rely on extractive models rooted in fear and hype. And it is not a coincidence, we would argue, that all of these predatory technologies take advantage of our students as both products and consumers as they seek to further monetize and capture educational experiences and intellectual property for the benefit of these companies rather than for teachers and students.
Given these realities, we argue for putting refusal back in-the-loop of the writing process, making opting-in an intentional and informed choice and not compulsory simply because Big Tech has told us it’s inevitable.
For us, refusal-in-the-loop writing expands on our work on Refusing GenAI in Writing Studies as it pushes for the ability to opt-out of corporate technologies—including but not limited to GenAI—rather than treating them as compulsory for 21st century writing processes. Refusal-in-the-loop writing also asks us to zoom out and examine the loop in full, looking beyond our individual screens and writing processes to grapple with the human and planetary costs of digital writing (Edwards, 2025). To do so, refusal-in-the-loop writing demands that we slow down and reflect on the broad implications of our responses to new technologies, including where and how we locate them within longer histories of technological development. For example, what are the implications of requiring or encouraging students to interact and experiment with GenAI in writing classrooms? How does compulsory adoption impact our capacities for academic freedom and ethical decision-making? Instead, refusal-in-the-loop writing considers a range of pedagogical possibilities for preparing students to write, think, and research.
Refusal-in-the-loop writing involves asking oneself why ‘AI’ is necessary for a particular writing task, and whether an analysis of its broad affordances, limitations, and impacts indicates that GenAI is a better option than consulting a person, library, search engine, or other writing resources. Refusal-in-the-loop writing also means refusing to repeat Big Tech marketing claims that do not include meaningful evidence, such as that ‘AI’ “will improve dramatically very soon,” or “take jobs.” In short, refusal-in-the-loop resists the language and logics of Big Tech as we consider what writing technologies we will use, when, and how we will use them.
AI hype and boosterism are predicated on easy compliance. Refusal—individually and collectively—matters. Through refusal-in-the-loop writing, we can continue to be the field that critically engages emergent digital technologies by leading with caution, care, and a clear-eyed understanding of our past.
Generative AI Disclosure Statement
No generative AI use to disclose.
References
Banks, A. J. (2006). Race, rhetoric, and technology: Searching for higher ground. Routledge.
Byrd, A. (2023). Truth-telling: Critical inquiries on LLMs and the corpus texts that train them. Composition studies, 51(1), 135–142.
DeVoss, D. (2019). TechnoFeminisms: A conversation about pasts, presents, and futures. Computers and Composition, 51, 68–78.
Edwards, D. W. (2020). Digital rhetoric on a damaged planet: Storying digital damage as inventive response to the Anthropocene. Rhetoric Review, 39(1), 59–72.
Edwards, D. (2025). Weathering the Rhetorical Climates of AI. College Composition & Communication, 77(1), 13–38.
Gerard, L. (2006). The evolution of the Computers and Writing Conference, the second decade. Computers and Composition, 23(2), 211–227.
Graham, S. S. (2023). Post-Process but Not Post-Writing: Large Language Models and a Future for Composition Pedagogy. Composition Studies, 51(1), 162–168.
Haas, A. M. (2007). Wampum as hypertext: An American Indian intellectual tradition of multimedia theory and practice. Studies in American Indian Literatures, 19(4), 77–100.
Haas, A. M. (2012). Race, rhetoric, and technology: A case study of decolonial technical communication theory, methodology, and pedagogy. Journal of Business and Technical Communication, 26(3), 277–310.
Haas, A. M. (2018). Toward a digital cultural rhetoric. In The Routledge handbook of digital writing and rhetoric (pp. 412–422). Routledge.
Hawisher, G. E., & Selfe, C. L. (1991). The rhetoric of technology and the electronic writing class. College Composition & Communication, 42(1), 55–65.
Jamieson, S. (2022). The AI “crisis” and a (re) turn to pedagogy. Composition Studies, 50(3), 153–157.
Janangelo, J. (1991). Technopower and technoppression: Some abuses of power and control in computer-assisted writing environments. Computers and Composition, 9(1), 47–64.
Johnson, G. P. (2023). Don’t act like you forgot: Approaching another literacy “crisis” by (re) considering what we know about teaching writing with and through technologies. Composition Studies, 51(1), 169–175.
Miller, C. R. (1979). A humanistic rationale for technical writing. College English, 40(6), 610–617.
MLA-CCCC Joint Task Force. (2023). Writing and AI Working Paper: Overview of the Issues, Statement of Principles, and Recommendations. Modern Language Association and Conference on College Composition and Communication.
Ridolfo, J., Hart-Davidson, W., & Lindgren, C. A. (2022). “Hello, is this the writing center?”: Illicit paper mill activity and the compromised recomposition of college and university websites. Kairos, 26(2).
Selfe, C. L. (1999). Technology and literacy: A story about the perils of not paying attention. College Composition & Communication, 50(3), 411–436.
Smalley, Suzanne. (2022). Balancing Student Privacy and Open Access. Inside Higher Ed. https://www.insidehighered.com/news/2022/02/15/course-hero-contends-student-privacy-concerns
Sparby, Derek M. “Digital social media and aggression: Memetic rhetoric in 4chan’s collective identity.” Computers and Composition 45 (2017): 85–97.
Stanton, C. (2023). A dis-facilitated call for more writing studies in the new AI landscape; or, finding our place among the chatbots. Composition Studies, 51(1), 182–186.
Trice, M., & Potts, L. (2018). Building dark patterns into platforms: How GamerGate perturbed Twitter’s user experience. Present Tense: A Journal of Rhetoric in Society, 6(3).
Vee, A., Laquintano, T., & Schnitzler, C. (2023). TextGenEd: Teaching with text generation technologies. The WAC Clearinghouse.
Vie, S. (2013). A pedagogy of resistance toward plagiarism detection technologies. Computers and Composition, 30(1), 3–15.
