When we observe the flurry of anxiety surrounding artificial intelligence in our staffrooms and leadership meetings, it is easy to assume that we are witnessing a technological crisis. AI is described as a fundamental threat to education, a cataclysmic change agent like nothing we have seen before. Teachers worry about authenticity, leaders debate policy, and schools scramble to install detector software in a frantic effort to safeguard academic integrity. Yet, as the dust begins to settle on this latest educational disruption, a quiet and uncomfortable truth emerges. Artificial intelligence is not the core problem facing modern education; rather, it is a mirror reflecting a vulnerability that has quietly persisted in our classrooms for decades.
AI makes it effortlessly simple for students to complete tasks, generate essays, and answer structured prompts. The speed with which students outsource their thinking to AI asks an uncomfortable question: did they ever see value in the work itself?
"The rise of artificial intelligence has not broken assessment; it has simply exposed how often we accepted the completion of tasks as a proxy for genuine intellectual engagement."
For years, school leaders and classroom teachers have relied on activities that act as convenient proxies for learning. We set a comprehension worksheet, request a standard five-paragraph essay, or design an end-of-unit quiz that asks students to recite dates, definitions, and formulas on demand. In each case, the completion of the artifact becomes the target. Yet, in many of these scenarios, there is remarkably little need for authentic cognitive struggle or deep conceptual change.
We have encountered this tension before. Two decades ago, educators grappled with the challenge of "Googleable" questions – queries that could be satisfied by typing key phrases into a search engine, required little to no thinking. Today, generative AI presents the same scenario dressed in a more sophisticated suit. If our assessments can be satisfied by an algorithm synthesizing internet data, then the issue does not lie solely with the algorithm – it lies in our reliance on tasks where execution can be separated from actual thinking (Ritchhart, 2015).
To understand why this distinction is critical, we need to contrast the typical compliance-driven classroom assignment with the rich learning that occurs when a person has a genuine need or desire to master something new.
Consider the learning that happens outside the formal structures of schooling. When an individual decides to repair an engine, master a musical instrument, build a piece of furniture, or understand a complex historical event out of sheer curiosity, the process looks entirely different. The acquisition of the new skill or knowledge carries high personal value. The learner is driven by an internal imperative: they need this capability to achieve a goal they care about, or they are captivated by the topic itself. In these moments, taking a shortcut makes no sense, because bypassing the learning process deprives the individual of the very competence they seek.
It is not surprising, then, that researchers Jal Mehta and Sarah Fine uncovered a striking pattern when studying American high schools for their landmark study, In Search of Deep Learning. As they observed classrooms across the country, they discovered that the most profound, transformative learning rarely took place within the core academic curriculum. Instead, it flourished in co-curricular spaces – the drama productions, debate clubs, newspaper rooms, robotics teams, and athletics programmes that students voluntarily elected to join (Mehta & Fine, 2019).
In these elective spaces, the traditional mechanics of schooling were inverted. Students were not performing tasks to collect an exit ticket, satisfy a rubric, or secure a grade; they were engaged in authentic domain practice because they cared deeply about the outcome. Autonomy was high, purpose was transparent, and mastery was the natural goal. Mehta and Fine (2019) observed that these co-curricular spaces systematically fostered what they identified as the powerful trifecta of deep learning: mastery, identity, and creativity. Because participation was voluntary and the standards of excellence were held by an authentic community of practice rather than dictated by a mark book, students willingly invested the hours of struggle required to achieve real competence.
This dynamic is deeply rooted in human motivation and is cogently explained by Self-Determination Theory, developed by psychologists Richard Ryan and Edward Deci. Ryan and Deci (2000, 2017) posit that human flourishing and high-quality learning depend on satisfying three basic psychological needs: autonomy (feeling a sense of agency and volition), competence (feeling capable and effective), and relatedness (feeling connected to others and a broader purpose).
When learning activities are designed primarily to monitor compliance, we operate almost exclusively within the realm of extrinsic motivation. Students complete the task not because they value the understanding it confers, but to obtain a reward – such as a grade or an accreditation – or to avoid a penalty. Under these conditions, intrinsic motivation diminishes. Ryan and Deci’s work demonstrates that when individuals are moved by external controls, their engagement tends to be superficial, and they naturally seek the most efficient path to satisfy the requirement.
Conversely, when learning environment design supports autonomy and personal relevance, students internalise the value of what they are doing. The target shifts from task completion to authentic mastery.
"When learning is merely a path to a grade, a mark, or an exit ticket, students will naturally take the most efficient shortcut available. But when learning possesses intrinsic value, mastery becomes the only goal that matters."
When we contrast these two positions, the current dilemma around AI becomes transparently clear. When education is framed primarily as a series of hurdles to clear – a path to a score, an ATAR, or a credential – and when the knowledge itself is not valued by the learner for its intrinsic utility or interest, taking shortcuts is a rational human response. Students are simply optimising for the system we have created. In the past, those shortcuts took the form of copied homework, rushed cramming, or surface-level summaries. When a student uses AI to generate an essay, are they doing anything fundamentally different to what the generation before did as they relied on CliffsNotes? AI has not created the impulse to take a shortcut; it has merely automated it.
When the learning itself has intrinsic value, however, the dynamic flips entirely. A student who genuinely wants to understand how an ecosystem balances, or how to craft a persuasive argument to address a community issue, does not want an AI to do the thinking for them. They might use AI as a dialogue partner, a tutor, or a brainstorming tool, but they will not surrender the cognitive heavy lifting, because they value the resulting competence. Mastery, not the artifact, remains the objective.
Banning AI or doubling down on surveillance technology will not resolve this structural issue. Prohibiting access to generative models in schools attempts to preserve a model of assessment that was already failing to guarantee deep learning. While education is viewed by students as an exit ticket – a necessary transaction to get to the next stage of life – they will continuously search for the path of least resistance.
AI is not making this path possible; it is simply highlighting what has always been fragile within our traditional pedagogical designs.
The task facing educators and school leaders is not to figure out how to outsmart generative AI, but how to redesign learning experiences so that they are worth doing. We are invited to move away from setting tasks that function as mere proxies for compliance, and move toward cultivating learning dispositions, student agency, and authentic inquiry. When we design learning environments that cultivate autonomy, foster competence, and connect knowledge to real-world purpose, we do not need to police the use of AI. We will have restored the intrinsic value of learning itself, ensuring that our students choose the rich, challenging journey of real understanding over the simple illusion of completion.
References
Deci, E. L., & Ryan, R. M. (2000). The "what" and "why" of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01
Mehta, J., & Fine, S. (2019). In search of deep learning: Inside the effort to remake the American high school. Harvard University Press.
Ritchhart, R. (2015). Creating cultures of thinking: The 8 forces we must master to truly transform our schools. Jossey-Bass.
Ryan, R. M., & Deci, E. L. (2017). Self-determination theory: Basic psychological needs in motivation, development, and wellness. The Guilford Press.