Bespoke or Prescribed? The Myth of Personalised Learning

Among the most frequently cited benefits of AI in education is the promise of personalised learning. Ask any AI chatbot, ChatGPT included, about AI’s role in the classroom, and personalised learning inevitably tops the list. But who created this narrative, and when did we collectively decide that this was the pinnacle of AI’s potential in education? Why is this particular vision so widely emphasised, and what assumptions underpin it?

At its core, personalised learning aims to tailor educational content, pace, and delivery to the individual learner. But this seemingly innocuous idea conceals a host of complex, and often unexamined, questions. Who decides what is personal? How does a machine determine what a learner needs? And crucially, what does “personalised” actually mean in practice? Is it about addressing a learner’s strengths and interests, or merely remediating their weaknesses as perceived by an algorithm?

Personalisation to What End?

There is an implicit goal in most personalised learning systems; a destination, a path, a defined outcome. But personalised to create what? Are children being guided towards a predefined set of competencies? Is there an end goal, and if so, who sets it? More provocatively, should there be an end goal at all? Education, at its best, is about exploration, freedom, and the messy, often non-linear journey of learning. Mistakes, detours, and curiosity are not inefficiencies to be optimised away, they are the substance of learning itself.

AI-driven personalisation often assumes that progress is measurable, that learning can be quantified and optimised. But what about the elements of progress that defy measurement? What about learning behaviours; resilience, adaptability, curiosity, that emerge over time and through experience? What about social skills, collaboration, and empathy? Personalised learning, by design, is an individualised experience. But education is, at its heart, a social endeavour. If personalisation leads to isolation, what kind of learners - and people- are we creating?

Moreover, when every learner is placed on their own optimised track, where is the shared experience of learning; the chance to debate, to disagree, to collaborate, and to co-construct knowledge with peers? The risk is that in chasing efficiency and personalisation, we erode the very experiences that prepare learners for the complexities of the world beyond the classroom.

Transparency, Agency, and Trust

Another pressing concern is transparency. How do personalised systems make decisions about what content a learner sees, what pathways they are offered, or what feedback they receive? These systems often function as black boxes, with little clarity on how recommendations are generated. If a system suggests that a learner revisit a concept or accelerate through material, on what basis is that judgement made? Who controls the algorithm’s parameters, and whose values are embedded in those decisions?

More importantly, does the learner have any agency in that decision? Personalised learning risks becoming a one-way street, where learners are guided, or nudged, by unseen hands, without awareness or consent. The irony of personalised learning is that it often sidelines the learner’s voice. Surely, the best person to decide what is best for you is you. Yet, in many AI-driven models, the learner becomes a passive recipient of algorithmically curated experiences, rather than an active participant in shaping their own learning journey.

There is also the issue of trust. For personalised learning systems to be effective, and ethical, learners and educators must trust them. That trust can only be built through transparency, accountability, and a genuine commitment to placing the learner’s needs, aspirations, and agency at the centre. Without these, personalisation becomes not a gift, but a constraint.

The Future of Learning

So, what should learning look like in the age of AI? It should be exploratory, not prescriptive. It should encourage freedom and creativity, not just efficiency and optimisation. There is a danger in allowing AI to guide learning along narrow, preordained paths. True personalisation is not about being guided through a system; it’s about having the autonomy to shape your own path.

Learning should be grounded in human experience; rich with dialogue, collaboration, unpredictability, and play. Personalisation, if it is to be meaningful, must recognise the full complexity of what it means to learn and to grow. It must make space for exploration, for mistakes, and for the kind of deep, personal engagement that cannot be captured by metrics or optimised by algorithms.

If we are to reclaim learning from the algorithms, we must ask harder questions about the narratives we have accepted, the systems we are building, and the learners we are shaping. AI should not dictate learning; it should support the learner’s agency, creativity, and humanity.