Evaluating Arguments About GenAI

Key Points:

Jennifer Sano-Franchini, West Virginia University ¹^,²

Maggie Fernandes, University of Arkansas
Megan McIntyre, University of Arkansas

We created Refusing Generative AI in Writing Studies in response to the persistent pattern of refusal perspectives being misrepresented as fear-based, irrational, and uninformed (Guglielmo; Kantrowitz; Marantz; Patel; Waite). We understand this rhetorical move to quickly dismiss refusal perspectives as a strategy that works to advance AI marketing, facilitate upward wealth redistribution, and silence opposition/dissension. Furthermore, the widespread misrepresentation of refusal as a position and strategy limits the ability of students and instructors—particularly those in precarious positions—to make choices about how to interact with emergent GenAI technologies. We wanted to state in clear and certain terms that there are sound reasons to be wary of and even to resist these technologies, as outlined in our Quickstart Guide.

Because so much of the conversation about GenAI in higher education is being driven by Big Tech and large, profit-driven EdTech companies, we need to engage information about GenAI and higher education in a critical and intentional way. Rhetorical awareness and analysis can help us differentiate between marketing tactics and the uptake of those tactics, informed opinions, and reasonable concerns about the utility and ethics of GenAI for writing and learning. As such, this page relies on longstanding disciplinary knowledges and perspectives in writing studies about rhetorical analysis and rhetorical contexts.

Below are questions and considerations for evaluating arguments that promote the adoption and use of GenAI products (while simultaneously denying opportunities for questioning, resistance, and refusal as informed responses) with a broad focus on analyzing flows of power.

Analyzing How Audiences Are Positioned

Readers will find that we focus primarily though not exclusively on arguments for GenAI adoption in education. We refuse to engage in bothsidesism as we believe it is important to recognize that the various sides of the argument are not all created equal—in other words, arguments for GenAI adoption often rely on unsubstantiated marketing claims, bolstered by the interests and capital of Big Tech and those who benefit financially from the success of these companies.

We believe an important part of critical digital cultural literacy is being able to critically assess the information presented to us about technologies and products online and offline. We hope that these considerations will help us all move to more nuanced understandings of not only whether, when, why, and how GenAI products should be adopted or used, but also how we might evaluate other arguments that are presented to us about a range of products and technologies online and offline.

What does the person making the argument have to gain from the position they are taking?

What is their professional and educational background (being careful not to engage in educational elitism)? Check their LinkedIn profile, and review their professional websites, resumes, and CVs. Does their title mention “AI,” “EdTech,” or similar terms? Do they work for companies that rely heavily on GenAI technologies? This is not to say that those in these positions cannot have opinions about GenAI, but rather that it is important to be aware of how those who work for companies like OpenAI and Grammarly are often engaging in marketing when they are talking about these products.

Have they openly disclosed any connections to GenAI companies, products, or integration projects? Where does that disclosure happen? Although those who have consulted for these companies might have unique insights into the inner workings of companies like OpenAI, these relationships can also inevitably shape one’s perception of these products. Disclosing connections and recognizing how these material relationships impact how one interacts with GenAI is important for transparent conversations especially as some may perceive the values and interests of companies like OpenAI to be at odds with the values and interests of higher education and writing instruction.

What are the various incentives and outcomes circulating in the arguments being made—not only for the individual, but also for institutions, corporations, investors, and other stakeholders?

How does the argument exist in the context of the author/speaker’s broader reputation, including their institutional position and/or social media and online presence?

Do they show a willingness to acknowledge legitimate and reasonable arguments for refusal?

Do they show an understanding of—or willingness to learn about—ongoing conversations in disciplines that focus on these intersections, including but not limited to computers and writing and digital rhetoric, media studies, and science and technology studies?

Do they demonstrate an awareness or willingness to learn about research about digital technologies, including GenAI, and their cultural, economic, ethical, and political implications?

What are they asking for from their audiences?

Do their goals include approaches that center the needs of GenAI companies or products and/or the purchase or adoption of a specific product?

Whom do they quote or cite frequently?

What is the credibility of those who are quoted or cited?

Does the source engage a range of perspectives from a variety of experts?

2. Analyzing the Outlet/Platform

Where is the argument being presented?

Is it in an academic or popular outlet? Is it provided by a for-profit media company? A public non-profit news outlet? Is it on social media?

What is the publication’s review process, if any?

How is the information checked for factual accuracy?

What is the ethos or reputation of the source?

Does it trend toward particular kinds of sociopolitical or cultural biases? Who benefits from those biases?

Does the source have a history of circulating misinformation?

Does the source engage in practices and conventions designed to uphold journalistic integrity?

How is the source funded?

Are there ads? Underwriting?

Does the hosting organization rely on donations?

Who are the various parties who profit or otherwise benefit from the publication source, and how do they benefit?

What steps or actions does the source encourage?

Who benefits from those actions?

3. Analyzing Common Arguments

Signs of GenAI Marketing

Personifies GenAI technologies as assistants or editors
Describes GenAI technologies/products/tools as magical
Repeats Big Tech talking points and claims about GenAI products
Makes grandiose claims about what these products will do in the future with little if any support for those claims
Assumes GenAI use as a given; presents GenAI adoption as the only viable option with little if any support for this perspective

Stock Arguments in Current GenAI Conversations

It’s inevitable, so we might as well get ahead of it by using it.

This argument denies that we have any agency in shaping conversations about GenAI in higher education and beyond. It also dismisses research that indicates GenAI technology may never improve and is increasingly insolvent as a business (Zitron).

You’re doing students—especially already-marginalized students—a disservice by refusing to teach them how to use GenAI effectively and responsibly.

This assumes that using GenAI teaches students—including marginalized students—competencies that can’t be taught more effectively by other means. Additionally, this argument dismisses the short- and longer-term impacts of homogenized language and algorithmic bias on marginalized students, some of whom are already being falsely accused of AI-based academic integrity breaches (Gerezgiher; Hale; Mathewson).

Some people can’t refuse, refusing is a privileged position.

If refusal is a privilege, we should be working to expand access so that all instructors and students can exercise refusal as a right. We would argue that those with the ability to refuse should use their privilege to normalize refusal and make it an option available to all.

GenAI can be useful in certain limited ways.

Refusal is about recognizing that we have choices related to the use of—and engagement with—GenAI. We find that GenAI is currently not more effective than the many existing resources or skills that writers can develop via engagement with writing as a social, reflective, iterative process that takes time and effort, and these skills don’t come with the social and environmental harms endemic to the use of GenAI.

AI is revolutionary in medicine (or another specific context) so GenAI is a necessary part of college curricula across the board.

This position erases differences between various AI-related technologies and conflates large language model-powered chatbots (or other GenAI products) with other algorithmic technologies and approaches that work differently, have different impacts, and whose return-on-investment calculations are different.

We can teach students about the ethical challenges of GenAI so that they become critical and responsible GenAI users.

Teaching students to use and rely on GenAI alongside lessons about the serious ethical problems of these products undermines the seriousness of these problems. It is not enough to acknowledge criticisms if we are teaching students to contribute to a damaging industry.

GenAI is just a tool, and we should think of it like any other writing tool—like a pen, or computer.

This claim relies on a false equivalency and ignores how GenAI differs from other educational technologies, namely how this technology has serious implications for the environment, information bias, etc. See also, Premise 6 from our Quickstart Guide.

We can address academic integrity issues with GenAI through the use of surveillance technologies.

These arguments focus primarily or exclusively on issues of plagiarism with surveillance technologies presented as the solution. These positions often ignore the many other, more devastating impacts of widespread uptake of these products, and ignore the harm that classroom surveillance technologies do to faculty-student relationships, student learning broadly, and student writing processes more specifically.

Logical Fallacies in Arguments for GenAI Adoption

Strawman arguments that misrepresent refusal perspectives, especially assumptions that refusers are not paying attention to GenAI developments or are unaware of how GenAI works.

False dichotomy, false dilemma, or either/or, reducing positions about GenAI as either for or against.

False equivalencies, or hasty generalization, e.g., other technologies are also bad for the environment, or humans consume just as much water as GenAI products do.

Slippery slope, e.g., if we don’t teach students how to use GenAI critically now, they will not be able to succeed in their careers.

For more about logical fallacies:

“Logical Fallacies” on the Purdue OWL
“Fallacies” on the Internet Encyclopedia of Philosophy
“Informal Fallacies,” by the Department of Philosophy at Texas State University

Identifying Taken-for-Granted Values and Assumptions Bolstering Arguments for GenAI Adoption

How does the argument position writing—what it is, what it is good for, its purposes and outcomes, etc.?
Are the claims informed by the values of capitalism, which assume that a company’s profit for its own sake is a legitimate motivation and priority over public goods, e.g., public health, addressing income inequality?
Does the argument rely on progress and efficiency narratives?
Does the argument rely on the normalization of free labor—the kinds of labor that companies typically pay for—to contribute to GenAI development and ultimately profit for Big Tech?
Does the argument take for granted that we need to follow the lead of technocrats who are in power, and not the other way around?
Does the argument rely on efficiency outcomes and the co-optation of DEI and accessibility discourses at the risk of increased labor expectations?

4. Analyzing How Audiences are Positioned

What kind of agency are users positioned to have within the argument?
What is marketed on the platform? What does it assume about the viewer’s interests and values?
What are the values embedded within the platform and what does it say about who its presumed audience is? For instance, does the platform or argument assume GenAI use? Does it include mostly celebratory representations of GenAI?

References

Gerezgiher, Feven. “‘A death penalty’: Ph.D. student says U of M expelled him over unfair AI allegation.” MPR News. 17 Jan 2025.

Guglielmo, Connie. “Silicon Valley Fights ‘Doomer’ Bill, Taylor Swift Actually Didn’t Endorse Trump.” CNET. 26 Aug 2024.

Hale, Rachel. “She lost her scholarshipover an AI allegation — and it impacted her mental health.” USA Today. 22 Jan 2025.

Horowitz, Juliana Menasce, Ruth Igielnik, Rakesh Kochhar. “1. Trends in income and wealth inequality.” Pew Research Center. 9 Jan 2020.

Kantrowitz, Alex. “The Claims That ‘A.I. Will Kill Us All’ Are Sounding Awfully Convenient.” Slate. 14 Nov 2023.

Marantz, Andrew. “Among the A.I. Doomsayers.” The New Yorker. 11 Mar 2024.

Mathewson, Tara García. “AI Detection Tools Falsely Accuse International Students of Cheating.” The Markup. 14 Aug 2023.

Patel, Nilay. “Barack Obama on AI, free speech, and the future of the internet.” The Verge. 7 Nov 2023.

Waite, Thom. “Doomer vs Accelerationist: the two tribes fighting for the future of AI.” Dazed. 24 Nov 2023.

Zitron, Edward. “OpenAI Is a Bad Business.” Where’s Your Ed At. 2 Oct 2024.

Endnotes

Recommended citation: Sano-Franchini, Jennifer, Maggie Fernandes, and Megan McIntyre. “Evaluating Arguments About GenAI.” Refusing Generative AI in Writing Studies. Feb. 2025. https://refusinggenai.wordpress.com/evaluating-arguments ↩︎
Thanks to Tara Salvati for copyediting this page. ↩︎

Summary:

Jennifer Sano-Franchini, West Virginia University 1,2Maggie Fernandes, University of ArkansasMegan McIntyre, University of Arkansas We created Refusing Generative AI in Writing Studies in response to the persistent pattern of refusal perspectives being misrepresented as fear-based, irrational, and uninformed (Guglielmo; Kantrowitz; Marantz; Patel; Waite). We understand this rhetorical move to quickly dismiss refusal perspectives as a strategy…

Source: Evaluating Arguments About GenAI