Opinion
Across the HE sector, we have been designing programs against the wrong standard.
Not through negligence, and not without good intentions;, but against an assumption about what universities are responsible for producing, that the regulatory frameworks have never actually required.
That assumption is now driving curriculum overload, assessment overload, and a great deal of the current panic about whether AI has made it impossible to know if graduates are capable. Getting clearer about what the standard actually is, changes everything about how we respond.
The Australian Qualifications Framework describes graduate capability at bachelor degree level as broad and coherent knowledge and skills for professional work and further learning. Not exhaustive professional knowledge, not comprehensive coverage of everything a profession requires. The Higher Education Standards Framework (Threshold Standards) 2021 requires that students demonstrate the learning outcomes specified for the program, not every possible outcome their profession might demand. These are threshold standards. They define a floor of readiness to enter practice, not a ceiling of professional independence.
The tension here has a specific policy history. When the former Coalition government introduced the Job-Ready Graduates package in 2020, then-Education Minister Dan Tehan was explicit: "Universities must teach Australians the skills needed to succeed in the jobs of the future." The package built a funding architecture around that premise, incentivising universities to produce graduates ready for immediate employment. The Australian Universities Accord softened the language but retained the underlying expectation: that universities are the primary site at which workforce capability is formed and delivered. The Accord has moved on from that framing. The idea hasn't.
The assumption that programs should produce complete professionals has a visible shape. Programs accumulate units to cover more ground, close more gaps, pre-empt more of what graduates might encounter in practice. Unit coordinators load tasks onto their twelve-week units because they feel personally responsible for certifying capability that the whole system seems to demand. The result is assessment that is simultaneously overloaded and underconfident: too much asked of each task, too little trust in the picture they collectively build. It is what happens when a program is designed to produce complete professionals rather than capable graduates ready to enter practice.
This is not a distant problem. For a regional university whose graduates often become the town's only nurse, only teacher, only social worker, the community expectation is concrete. They expect someone who can do the job. That expectation is legitimate.
The question is not whether to honour it, but how. Loading more onto every unit and every assessment task is not the answer. The answer is to design at the program level, around what can actually be assured at graduation, and to let that design determine what each unit needs to do, rather than asking each unit to carry the whole burden alone.
A program can build a coherent, evidenced picture of threshold capability. It cannot, and should not, try to replicate the full complexity of professional practice. That work belongs to the next phase.
What the current AI moment has done is not lower the standard of learning we expect. It has made the evidence weaker. But AI is not the source of this problem.
Contract cheating, essay mills, and the quiet art of strategic surface learning have been undermining the evidence base for years. What AI has done is make the problem impossible to ignore.
Two TEQSA-published resources on assessment reform, the most recent from 2025, make the point plainly: detecting AI use in submitted work is, for practical purposes, impossible. The result is that the products we have traditionally relied on as proxies for learning: essays, reports, take-home tasks, are now less reliable indicators of what a student actually knows and can do. The response is not primarily about detection. It is about redesigning assessment to generate trustworthy judgement through multiple, contextualised approaches gathered across a program, not crammed into each unit.
That evidence is not a better version of what we already collect. It is gathered differently, across time, through observation of performance, through interaction and demonstration, through the professional judgement of people who have watched a student work. Work integrated learning, when deliberately assessed rather than treated as experience alone, offers some of the richest evidence of threshold capability available to a program. Programs in medicine, nursing and allied health have long understood this: a medical student's readiness for practice is built from direct observation of clinical encounters, structured feedback over placements, and the professional judgement of experienced practitioners. Short oral assessments function as verification points that are genuinely difficult to fabricate. Together, they build a picture that a single submission never could.
The professions themselves have always assumed that this picture would be incomplete at graduation. Medical graduates enter supervised internship before they practise independently. Psychologists undertake extended supervised practice before full registration. Teachers move from provisional to full registration through mentored classroom experience.
Even in less formally regulated fields, graduates typically enter roles with reduced responsibility and close oversight, with an explicit expectation of continued development.
What emerges is not a gap between higher education and practice, but a handover. Universities establish that a graduate has reached a threshold of capability, sufficient to enter practice, continue developing, and work safely under appropriate oversight. What graduates do not arrive with is complete professional knowledge, full independence, or readiness for every situation their field will present. That is not a failure of the program. It is the design. Industry and the professions then take up that work, not as a fallback for what universities failed to finish, but as the designed site where capability is extended, tested and made fully visible in context. That requires employers and professional bodies to take their role in this seriously: structured induction, supervised practice and graduated responsibility are not optional extras. They are the other half of the system. Where that structured entry is absent, the gap between university and practice becomes exactly the problem everyone fears.
We have been holding ourselves to a standard the system never required. A degree is not a certificate of professional completion. It is evidence that a graduate has reached the threshold, ready to enter practice, continue developing, and take up the work that only practice can finish.
As more institutions move toward programmatic assessment, the design decisions being made right now will shape what evidence universities can build and what claims they can defensibly make at graduation. This is the moment to get the standard right. Designing programs around threshold capability, rather than the assumption that universities must produce complete professionals, is the most important curriculum decision leaders can make right now.
Associate Professor Prue Laidlaw is Academic Director, Education Strategy at Charles Sturt University.