Australia has no shortage of advice about artificial intelligence in classrooms and lecture theatres. What it lacks, according to a pointed new analysis, is a shared destination. The country has been handed a compass when what learners, teachers and employers actually need is a map.
That is the argument at the heart of an analysis published by Campus Review, which contends that despite genuine strengths in pockets of the system, Australia’s overall response to generative AI in education and training is fragmented, stalled, and increasingly at odds with the nation’s own ambitions on social cohesion and economic productivity. The piece frames the problem as one of coordination rather than capability: the ingredients exist, but they are scattered across sectors that rarely speak to one another.
The news
The core claim is that generative AI has moved through Australian education faster than the institutions meant to govern it. Schools, universities and the vocational education and training sector have each produced their own guidance, principles and cautious position statements, yet those documents point in different directions and rarely connect. A student who learns one set of expectations about AI use in Year 12 can arrive at university to find another, then step into a traineeship or a job governed by rules that resemble neither.
The analysis argues these are not minor irritations but systemic risks. When learners transition into, between and beyond formal education, the gaps in policy and institutional action leave them without consistent signals about what responsible AI use looks like. That inconsistency, the piece suggests, undermines the very outcomes governments say they want from AI, namely a more productive economy and a more cohesive society. Guidance that tells everyone to proceed carefully, without agreeing on where they are heading, functions as a compass and nothing more.
The timing matters. Australian institutions are already spending heavily on AI systems and skills, and universities in particular have been pouring money into new tools while acknowledging significant capability gaps among staff. Investment without coordination, the argument runs, risks buying a great deal of technology that no two parts of the system use in the same way.
Two ways of reading the problem
There is a case for viewing the current patchwork as a feature rather than a failure. Education in Australia is deliberately federated, with states running schools, the Commonwealth shaping higher education funding, and a separate architecture governing VET. Advocates of local autonomy argue that a fast-moving technology is better handled by institutions close to their students than by a single national edict that could be outdated before it is signed. On this view, a compass is exactly right for terrain that keeps shifting, because a fixed map would send everyone confidently in the wrong direction.
The counter-argument, and the one the Campus Review analysis leans towards, is that autonomy has curdled into drift. Letting every sector improvise has produced duplication, contradiction and long stretches where nobody acts at all, because responsibility sits everywhere and therefore nowhere. Learners bear the cost of that stalemate. A shared framework, in this reading, need not dictate every classroom decision; it simply has to agree on the destination so that individual institutions can chart their own routes without contradicting each other at the boundaries.
What it means for Australia
The stakes here are unusually concrete for Australia because education is one of the country’s largest export earners and a central plank of its productivity agenda. International students choose Australian institutions partly on the strength and clarity of the qualifications they earn. If employers cannot trust what an AI-assisted degree or certificate actually certifies, the value of the credential erodes, and with it a major source of national income.
There is a workforce dimension too. Governments and industry groups have repeatedly named workforce capability as the decisive factor in whether Australia captures value from AI or merely consumes tools built elsewhere. That capability is forged in exactly the transitions the analysis worries about, from school to further study, from study to work, and from one job to the next as tasks are reshaped by automation. If the education system cannot give people a consistent grounding in how to use these tools well, the productivity gains ministers keep promising will be harder to realise, and the social divides that AI can widen will be harder to close.
The equity question sharpens the point. Well-resourced schools and universities can afford to build their own thoughtful AI policies, run staff training and buy enterprise-grade tools. Smaller regional providers and stretched VET colleges often cannot. A fragmented national response quietly advantages institutions that already have the most, leaving others to cobble together guidance from whatever is freely available. A shared map, by contrast, would give every provider a common baseline to work from regardless of budget.
What’s next
The analysis stops short of prescribing a single body or document to fix the problem, and that reticence is telling. The hard part of moving from compass to map in Australian education is not intellectual but political, because it requires the Commonwealth, the states and the VET sector to agree on shared goals in an area where each guards its own patch. Any credible next step would need to reconcile that federated structure with the reality that learners do not experience education in neat jurisdictional boxes.
Practical progress is likely to come first from the sectors themselves rather than from a grand national settlement. Universities are already comparing notes on AI-ready systems and skills gaps, and there is room for schools, higher education and VET to align at least on the transition points where students move between them. Agreement on what responsible AI use means at the end of school and the start of a degree, or between a qualification and a first job, would deliver much of the benefit of a national framework without waiting for one.
For now, the warning is that Australia is spending, deliberating and publishing at pace while the pieces refuse to fit together. The country has been navigating generative AI in education by feel, confident of the direction but unsure of the destination. Whether that is enough depends on how quickly a compass can be turned into something learners can actually follow.
Sources: Campus Review.
