September 15th, 2025

[ For more on The AI Hype Wall of Shame and our rating system, see this link.]

Rating: #twomarvins (a depressing excuse for journalism, with occasional inducements of paranoia)

UCI critical media theorist Dr. Ricky D. Crano catches Inside Higher Ed in the act of ethics washing and genuflection to ed tech.

Joining me is Jenny Billings, the program chair of Associate in Arts, and division chair of English, and study skills at Rowan-Cabarrus Community College in North Carolina. She’s seen improvements in retention and confidence… among students in the Developmental English Program at Rowan-Cabarrus after they started using AI. …One of those AI tools is Grammarly. So also joining us today is Jenny Maxwell, the head of Higher Education at Grammarly.

Inside Higher Ed’s latest podcast episode opens on a note so seemingly innocuous it almost passes unnoticed: the host introduces the guests and, with a breezy so, folds a Grammarly rep into the conversation as if a vendor’s presence were exactly the kind of thing one might expect in a candid discussion of using generative AI in higher education. The fact that the featured guest seems to be running or chairing at least three distinct programs out of a non-unionized, non-tenured post at a public community college in a notoriously underfunded state system goes entirely without comment. From this moment, the podcast’s framework is clear: not simply an infomercial but also the normalization of platform capitalism in education as the background condition for any serious conversation about student writing in Fall 2025. Here we are, not even three years out from the introduction of ChatGPT.

Strong writing is the foundation of academic achievement, but students often struggle with clarity, structure and confidence. With Grammarly, your students and faculty can access AI-powered writing assistants, helping them communicate more effectively, improve academic integrity and enhance student success.

Read by the host, this opening plug slyly positions Grammarly as the main AI player on campus; it’s the only tool name-dropped in the entire episode (a single mention of ChatGPT comes off so hastily that Apple’s automatic transcription tool rendered it as “catchie PT” 😂). As is common of the hype in the world of gen AI tooldom, Grammarly is framed as a corrective to human deficiency and a panacea for supposed undergrad fears. The appearance is that each student is now backed by a team of helpful assistants who just need some effective ordering around. At no point does any discussant think it worth asking: Who, really, is taking orders from whom?

Here’s what the podcast never says directly: AI-powered Grammarly positions students as if they are cogs in a corporate enterprise who can now outsource their tedium and automate the legwork, while (we must suppose) miraculously emerging with increased confidence. Like much in the world of tech solutionism, most of this supposed learning process is wholly logistical.

As its praises are sung, Grammarly is named more than thirty times, each repetition further eroding the distance between a serious discussion about teaching students and an effusive marketing campaign for a laggard in the race to win institutional contracts and student users.

Some background you won’t learn from listening to this podcast: Founded as a tool to make rapid-fire email communications less burdensome, Grammarly scaled through freemium and subscription models, cutting deals with university bureaucrats to begin quietly extracting rents with little to no faculty approval or oversight. In 2023, it partnered with OpenAI to integrate a GPT-3-based model trained on business writing, and in 2024 it launched its now flagship product, “Authorship,” an AI-assisted detection and feedback tool designed for integration into learning management systems like Canvas. When it acquired the productivity startup Coda at the end of 2024, Grammarly’s modus operandi was a fait accompli: a supposed labor-saving device that could be packaged as a win for efficiency, which in today’s academic marketplace is (supposedly) all that matters for student success. By 2025, Grammarly could claim forty million global users, adoption by tens of thousands of institutions, and awards for “ed-tech innovation.”

Inside Higher Ed’s “from hush hush to all in” podcast thus sets out to plop a cherry on top of the undermining of anything so quaint as faculty governance. In the real world you might be reading this post because you actually like to know what you are doing when you are in a classroom, but here’s a different way of looking at things.

[Faculty] don’t take semesters off to go research how to incorporate AI in their courses that they’ve been teaching for many years, right? So they had to fly the rocket ship while they’re building it, and the galaxy’s on fire.

Fear-mongering at this level pushes us briefly into three-Marvin territory (full-on paranoia). You’re not teaching any more: you’re piloting a rocket ship in a burning galaxy 🙄. Of course back on Earth, what’s actually burning is the planet, increasingly with the help of the massive energy costs of AI.

Like the conceit of “cloud” services, gen AI tools such as Grammarly like to portray themselves as metaphysical or weightless agents. In fact a single ChatGPT-style query consumes six to thirty times the energy of a conventional Google search (even though Google’s AI-overviews mean that energy-conscious users need to type “-ai” at the end of their search query to omit this feature).

But indifference to real-world environmental crisis is only one of the ways that Inside Higher Ed’s rampant product promotion ignores  material costs. For example, good luck getting the provost or IT department at your institution to reveal the rate it is paying for each student or staff member eligible to use the tool. Nor will you find much attention here to the persistent problem of errors, biases, confabulations, and faulty references. 

Instead, determined to take their listeners “all in,” the discussants in this podcast are selling efficiency,  a term they bandy as a description of student time saved or faculty labor reduced. That students may be learning nothing at all when they hand in autogenerated content does not seem to figure in the preferred metric.

[Y]ou know, you can use it for creation, you can use it for outlining, you can use it for revision, but we don’t want it to replace you and to replace your unique voice.

So use it to help you save time, use it to help you save energy and things that don’t really require that of you in the beginning, but we still want the unique student experience and the student voice.

Have your cake and eat it too much?  With “creation,” “outlining,” and “revision” all automated, what’s left to replace? And with AI “feedback” demonstrably ranging between sycophancy and uniformity, what sort of “unique” student voice are we expecting to find here?

Let’s be clear: Inside Higher Ed may want to “talk the talk” but what they are selling is writing instruction as a workflow: text is uploaded, processed, flagged, and returned; revision is algorithmically guided rather than dialogically negotiated.

At its most disingenuous, the podcasters talk about Marlee, a student who was falsely flagged by an AI detector, who lost her scholarship and faced mental health consequences. Although “we did not create the problem” of faulty AI detection, the Grammarly vendor intones, “we felt again, our moral obligation…to help our customer.” 

Clearly, what Marlee needs is not help from Grammarly but a teacher committed to sharing critical AI literacies. A teacher who knows that more technology is not the answer to a problem that this very technology has created. A teacher less focused on “detecting” crimes than on building student knowledge and judgment.

Of course, that kind of  decommodified pedagogy won’t help out Grammarly deliver return-on-investment for its $1 billion splurge into venture capital. So let’s put our money on the company’s “Authorship” dashboard which purports to “solve” the problem of AI-detection failure by—brace yourself—encouraging students to review their detection scores prior to submitting the assignment. This next step towards total AI integration, pits sly students against skeptical faculty in a cat-and-mouse game of techno-determinist oneupmanship. And the whole charade is paid for by institutions and/or their students!

Incredibly, the conversants appear to believe that everything that’s wrong with this picture can be saved by strategic use of the word alignment.

[I]t all comes back to aligning, do you believe that your role as an educator and my role as a vendor, product provider is to really enhance the students’ experiences?

Well, yes, but I don’t see how my role in enhancing a student’s experience can be boiled down to my belief in “align[ing] to the mission” of my institution—as if “alignment” were some kind of Hippocratic oath or sacrament.

…I think what we’re going to continue to see is an incredible moment where students are really going to drive behavior change at institutions, and they’re going to do that because they are the ultimate consumer of this experience.

As if students themselves are the real power drivers behind Grammarly! Raise your hand if the neoliberal notion of students as “consumers” of “experience,” um, aligns, with your ideas of learning. Or if your own vision of teaching “aligns” the kind of “success” that commercial ed tech can measure through a handy corporate dashboard. Efficiency, standardization, and quantified risk management are the order of the day! Pedagogy, imagination, or experimention not so much. “Education” is now a single trackable workflow. The old joke about the university being a hedge fund with a few classrooms attached has never felt closer to the truth.

According to the host, some institutions “just jump headfirst” into commercial AI, others are “morally” compelled to catch up, while a minority hope it will disappear.

I wanted to put some statistics in front of you from Inside Higher Ed‘s recent survey…

…the power of students saying, I know this technology is going to be a big part of the world of work for me, and I expect to be able to use this in my high-stakes experiences.

As distinct from student “power,” what the surveys actually demonstrated was a difference in uptake and interest that varies across gender, class, and first-generation status. Groups we might associate with the greatest exposure and responsiveness to AI hype (boys, white-collar workers, those with college-educated parents) are predictably more enthusiastic about AI potential.

It’s likely that none of these conversants have thought a whit about “aligning” their teaching with recent findings such as Komyna et. al.’s, out of the MIT Media Lab, underscoring the perils of “cognitive debt” in comparisons between “brain-only” writers and writers on autopilot.  Instead:

[H]ave you found that students are using it in a more, I guess, ethical way or more productive way than before?

By the close of the episode, that conclusion is foregone. Ethics is efficiency. Writing is output. Feedback is quantified and its costs carbon-free. The growing body of research showing actual decreases in worker productivity when AI is brought onboard (such as this one from the Upwork Research Institute or this one focused on coding work by the AI benchmarking nonprofit METR) suggests some combination of wild gullibility and demented data lust on the part of all involved.

[I]t’s gonna need to be included in some capacity in all curriculums, because it’s here.

“Because” here is a sad rhetorical device: a puerile tautology that my own college writers could mercilessly dissect. Like the casual so that opens the episode, the “all in” podcast closes with a chaser of heavy technodeterministic inevitablism.

What are we to make of this travesty? Supposedly Inside Higher Ed prides itself as “a fiercely independent voice” for “thoughtful, substantive analysis” on the most “pressing issues.” Yet in this podcast, it does not recognize the difference between journalism and advertising.

If Marvin, the Paranoid Android could cancel his subscription, I’m sure that he would.