14/04/2026
AI in Higher Education: Assessment at risk or curriculum rethink needed?
Contributor- Dr.Cristina Costa
Dr Cristina Costa is an associate professor and co-director of the MA in research methods in the Durham Research Methods Centre, School of Education, at Durham University in the United Kingdom.)
In the frenzied attempt to address the impact of generative artificial intelligence (Gen-AI) in higher education, universities have narrowed their concerns to a key question: how to design ‘secure’ assessments that resist the influence of Gen-AI large language models. Such framing implicitly casts students as being at risk of Gen-AI influence.
This is evident in institutional memos, working groups and practitioner discussions, which reflect a persistent concern with upholding academic integrity and ensuring that student work reflects their own effort. Yet, conversely, these same institutions champion Gen-AI as a tool for innovation and learning. These inconsistencies are too apparent to be ignored.
This contradiction is not just ironic; it is symptomatic of a more profound issue. Gen-AI has come to expose higher education’s long-standing curriculum crisis: the transactional, performative approach to education that has undermined the deeper aims of higher learning, that of fostering meaningful intellectual growth and critical inquiry.
In this vein, the arrival of Gen-AI is less a disruption than a revelation, exposing the systemic flaws that have shaped education into a transactional enterprise. The real crisis lies not in students having access to Gen-AI, but in the technocratic logic embedded in our curricula, pedagogical approaches and assessment regimes.
Instead of doubling down on control, what if we seized this moment to reimagine education as a collaborative, contextual and genuinely meaningful experience?
This reimagining must go beyond the concerns of academic integrity to confront fundamental questions about the purpose, meaning and practices of pedagogical work in higher education as both a space of higher intellectual engagement and meaningful practices.
Across higher education and society more generally, a revealing paradox is thus unfolding. This paradox stems from a broader technocratic ideology that dominates educational discourse. Gen-AI is treated not as a phenomenon that challenges existing systems, but as a resource to be controlled.
The result is a fragmented, even contradictory approach: universities rush to set student Gen-AI use within certain parameters while aiming to embed the same tools into administrative and academic workflows (for example, meeting minutes, etcetera). These inconsistencies create fissures in what is considered legitimate use, largely based on one’s institutional role.
As a result, students are caught in a conundrum: they are cautioned against unlimited Gen-AI use for their academic work yet are encouraged to adopt it to boost productivity and future employability.
This dual message casts students as both digital champions and potential cheaters, revealing deeper, unresolved tensions regarding the roles of education and Gen-AI in both knowledge and creative work.
Learning as performance
But the deeper issue is not student dishonesty, although is highly problematic to the development of a competent labour market. The key problem is the fact that a system that treats learning as performance – something to be measured, controlled and ranked – creates an artificial experience of learning, one that lacks genuine intellectual engagement.
If students are turning to Gen-AI to write essays, perhaps we should ask what this data reveals about the authenticity, relevance and meaning of the learning experiences we offer.
This question is crucial because Gen-AI offers a seductive promise: the outsourcing of intellectual effort.
Tech companies market these tools as enablers of personalisation, creativity and learner autonomy. Yet this promise often conceals a deeper conformity, tethered to logics of efficiency, optimisation and convenience. Notably, pedagogy rarely features in these narratives of technological progress.
This offers a version of ‘freedom’ that is misleading, to say the least. Students are told they can learn in new ways and based on their specific needs. Yet, this level of personalisation seems mostly to free them from the difficult yet necessary labour-intensive activity of thinking, so essential to opinion formation.
Framing Gen-AI as a solution for content delivery or cognitive automation risks reinforcing a ‘banking model’ approach of learning where knowledge is deposited rather than constructed. This sets the stage for the imposition of dominant ideologies and infocracy.
Misunderstanding critical thinking
Compounding this threat is a widespread misunderstanding of critical thinking. In many policy and institutional documents, criticality is conflated with problem-solving rather than problem-posing.
Yet, in a critical fashion, Gen-AI should prompt us to ask deeper questions about who it (dis)(em)powers, how and why. Merely asking what it can do leads to a rhetoric of convenience, not intellectual liberation.
One of the most urgent concerns is its potential to domesticate the mind, encouraging the reproduction of knowledge rather than its interrogation. Ironically, institutions respond to this threat by doubling down on ‘secure assessments’, acknowledging the risks but ignoring the broader implications.
Such responses reinforce a curriculum architecture where knowledge is delivered, reproduced and assessed in discrete, measurable units, and which is not open to contestation. Yet, what Gen-AI does very clearly is bring this structure into question.
If a machine can generate a passable essay in seconds, perhaps the assignment never truly required deep learning in the first place. Gen-AI therefore offers an opportunity to rethink what it means to learn, teach and work together.
But putting such an approach into practice is not easy. It requires confronting uncomfortable truths about our institutions, our practices and also ourselves. It will demand imagination, courage and, above all, a willingness to move beyond the rhetoric of crisis management and towards a vision of education that is humane, critical and just.
Thus, rather than designing AI-proof assessments, we might instead reimagine the curriculum itself, centring student engagement, creation and relevance applied to real-world contexts from which evidence of learning can stem more naturally. This is not about abandoning academic rigour, but it is about recognising that the most powerful learning is dialogic (with oneself as well as others), situated and relational.
Learning happens not only in tasks completed but also in the processes of forming relationships, challenging perspectives, and engaging with new practices.
Gen-AI, with its uncanny mimicry of human language, presents a unique opportunity to recentre these human elements, not as a patchwork exercise, but as a foundation to a higher education experience.
One of the most troubling effects of the Gen-AI panic is the devaluation of teaching itself. If content can be generated instantly, what value do classes have? If essays can be produced with a prompt, why engage in study?
But this line of questioning misunderstands the value of education. Teaching is not about information transmission; it is about shared intellectual encounters with complexity and uncertainty and the meaning one can extract from it. Learning is an engaged practice that cannot be outsourced.
Reclaiming learning
Additionally, Gen-AI cannot replicate the emotional and intellectual dynamics of a classroom discussion, the vulnerability of grappling with complex ideas, or the recognition of having created something meaningful.
To reclaim learning and teaching as relational practices is thus to resist the logic of efficiency and embrace the messiness of learning. It means designing learning not targeted solely at task completion but as a shared endeavour rooted in curiosity, mutual respect and transformative potential.
This also implies rethinking how to structure academic work, and teaching and learning time, as well as how to involve students in their own and others’ learning.
The institutional turn to surveillance in response to Gen-AI reflects a broader shift in higher education toward risk management, performativity and accountability. These trends create cultures of distrust and also compliance by the different parties involved.
The current discourse around Gen-AI in higher education is, at best, a missed opportunity. At worst, it risks reinforcing the very problems it promises to address. By focusing narrowly on assessment integrity, universities fail to engage with the deeper pedagogical and philosophical questions Gen-AI raises.
But there is another way. We can treat this moment not as a threat but as an invitation to rethink curricula, to reclaim the relational core of learning and teaching, and to rebuild education on the foundations of trust, critical dialogue and shared purpose.
A pedagogy of solidarity
What if, instead, we imagined a pedagogy of solidarity? One in which students are treated as engaged citizens, invested in learning how to contribute meaningfully to society; as an approach that prioritises trust, responsibility, care and the celebration of genuine learning.
Such a shift requires more than new policies. It demands a rethinking of institutional ethos and the curriculum, not only from institutions but also from academic staff as well as students.
This would mean investing in staff, prioritising learning participation over information consumption, embedding lived experiences in curriculum design, and valuing assessment as a process, not just as a product.
Instead of regarding Gen-AI as the enemy, we can choose to see it as a welcoming provocation that calls for deep reflection as well as bold action.
Ultimately, the debate around Gen-AI is not about technology; it is about purpose and meaning. What is higher education for in an age of machine-generated content? If its aim is merely to credentialise and prepare students as compliant citizens in and for the labour market, then the turn toward automation is unsurprising.
But if education aspires to cultivate critical, autonomous and creative citizens, we must resist the technocratic imposition and recover education’s emancipatory potential. It is high time we claim what we stand for!
