Contradictory Resistance Amongst Students
This blog post by Dylan Orchard, part of the AI and the Digitalisation of Education blog series, challenges the utopian technosolutionalist narratives paid for by AI corporations, and invites readers into a conversation on alternatives.
‘Artificial Intelligence is in education and it’s not going anywhere’ – that’s the almost universal sentiment around higher education at this time. Something inspired varyingly by Fear of Missing Out (FOMO), cost cutting, tech company lobbying and government enthusiasm. Something which is also often as ignorant of potential harms and effects as it is enamoured by the fantastic promises of tenuously defined ‘AI’ technologies.
Rejecting that narrative of inevitability (Guest et al. 2026) is no bad thing and burgeoning critical and Luddite voices are slowly emerging to do just that, although rarely loudly enough to reach the ears of policy makers either nationally or at institutional levels. At least not in the UK, where our government has entered a dizzying AI hype frenzy. An increasing amount of research, interest and organising energy is however being turned towards questioning the frenzied push for AI in education. Amidst that growing network of healthy cynicism though there’s one element that seems as inevitable and unquestioned as ever – students will use it.
Students will use AI, or more specifically LLMs, to cheat, summarise, translate, manage workloads, skip lectures and a dozen other things. This, the tech companies will tell you (as they push free subscriptions on undergrads), is because the ‘tools’ are revolutionary, the way of the future and too good to turn down. That hype narrative, that avid technosolutionism (Morozov 2013) to the exclusion of pretty much all else is an issue in itself; but the assumption of certainty that’s been framed by those arguments is, on a mundane level, something that can be interrogated more immediately.
For a long time education has been drifting towards a logic of the audit (Madianou 2024), a commercialised fixation on the output of ‘positive’ numbers, be it exam results, graduate numbers or more abstracted metrics designed to symbolise the whirring productivity of a machine whose purpose has become ever vaguer amidst the din of its own functioning. When positive use cases are suggested for students within that system one of the first questions asked should really be ‘positive for what’?
In my own, limited, experience as a teacher—I’m only a PhD researcher now myself, doing GTA work on the side—I’ve already heard numerous contradictory takes around that question. I’ve come across students who are entirely enthused about the potentials of LLMs in their education, convinced of the value of them even as I can’t help but disagree with their eager adoption. I’ve also, however, met students whose use of them is framed more by resignation than purpose—students who feel overloaded, who don’t see the value of specific work, who have been left unprepared for the environment they’ve been introduced to and the work it requires of them. You can take those perspectives as valid or invalid, excuses for seeking the AI workaround to reading lists or essay requirements or genuine critiques of the educational structure they’ve been introduced to. Either way though, you have to engage with their perspectives to judge and having judged, if they’re right about the inefficiencies of the system they’ve been pushed into; you have to ask whether the LLM is the solution or just the solution left to them by broader failures.
The tech companies don’t care either way, in their desperate pursuit of a user base and use cases: what better options there may be is an irrelevant thought, unless those options involve them even further. All too often it seems to be an irrelevant question to institutions too – where the requirement to mimic business and the logic of the audit ‘AI’ can pander to is answer enough. And that leaves educators and students alone, at all levels, to explore the issue.
When it comes to teaching, assessment, grading, the broad cultural and social realm of higher education – is everything as good as it can be? Are LLMs placed on top of a functional landscape as a genuine improvement rather than another commercialised fix? I don’t think anyone could, with a straight face, say the former is true. Certainly in the UK the impacts of austerity, mismanagement, the political appropriation of education as number generator, anti-intellectualism and a lot of other things have long driven the educator on the ground into a defensive position, navigating broken machinery in search of positive outcomes just as students are now doing with the fix of LLMs.
That’s a professional, sectoral and personal challenge for all of us. To change what’s broken is a big ask certainly and one with a lot of different answers. As far as the immediate situation goes though i.e. the direct engagement with students in the age of LLMs, we can look to their actions to outline what broken part may be doing the most harm and, perhaps, mediate it somewhat. Perhaps certain things are within our power to change, or at the very least consider and, if nothing else at all, they at least present points where we can be honest both with students and ourselves.
We can acknowledge that what they’re pushed to do is not always the best path for their education; we can acknowledge that they may need to navigate the higher education machine as best they can; and most of all we can make it clear that sometimes the tools they feel pressured to turn to for that aren’t good even if they may seem necessary. We can outline harms and acknowledge possible futures that may negate them, ones that assert themselves before relying on the technological fix, ones that come from those we teach as well as those we teach with. If we can all at least come to the point of challenging the narrative of utopian, technosolutionist inevitability then thinking about the next step may become, if not clearer, then one we might take together.
References:
Guest, O., Suarez, M., Müller, B., van Meerkerk, E., Oude Groote Beverborg, A., de Haan, R., Reyes Elizondo, A., Blokpoel, M., Scharfenberg, N., Kleinherenbrink, A., Camerino, I., Woensdregt, M., Monett, D., Brown, J., Avraamidou, L., Alenda-Demoutiez, J., Hermans, F., & van Rooij, I. (2026). Against the Uncritical Adoption of ‘AI’ Technologies in Academia. Zenodo. https://doi.org/10.5281/zenodo.17065099
Madianou, M. (2024). Technocolonialism: When technology for good is harmful. John Wiley & Sons.
Morozov, E. (2013). To save everything, click here: the folly of technological solutionism. First edition. PublicAffairs.
Further Reading
Crano, R. D. (2025). A Pedagogy of the Inevitable. Critical AI, 3(2). https://doi.org/10.1215/2834703X-12095991
Author
Dylan Orchard is a PhD researcher at King’s College London focusing on everyday resistance to AI and its motivations. He also works as a Graduate Teaching Assistant at KCL as well as having taught on courses around critical AI use for younger age groups. As well as resistance, his academic work focuses on faux-social LLMs and destituent educational models.