Published in The Australian, 25 May 2026
View source (subscription required)
A year 12 biology teacher asks her class a question about the role of vaccinations in providing herd immunity. Several students glance down, type it into AI and one reads back the answer. No pause. No process. No learning.
A history teacher assigns his year 9 class an essay on major social shifts during the Renaissance. Some students work hard, consulting a range of historical sources. Others ask AI to write their essay for them.
The essays written by AI effectively summarise large amounts of information. But research is clear that learning is damaged when students rely on AI to do their work. By contrast, students who do the work themselves are more likely to learn more – even if their essays get lower marks.
These scenarios are not fringe cases. New research shows AI has moved into schools much faster than have plans to deal with its consequences. Released on Monday, Learning First’s report, AI use in schools: Taking action now, draws from surveys, focus groups and interviews of more than 4000 NSW teachers and school leaders in each sector.
Time and again, teachers spoke of students using AI to get an answer the fastest, most convenient way possible. For these students, AI is not a partner in learning but a surrogate to which they delegate their thinking, a phenomenon the research has framed as cognitive outsourcing.
About half of secondary teachers surveyed say students are using AI for schoolwork. But figures for student use are almost certainly underestimates, since teachers say AI use can be hard to detect. Of these secondary teachers, about three-quarters say students use AI to complete assessments, despite the fact that more than 80 per cent say students face restrictions on using AI for this purpose. When students use AI to think for them, they short-circuit the learning process.
AI is typically designed to be maximally helpful and minimise cognitive effort. Yet productive cognitive effort is the very thing needed for learning to occur. Students who don’t know much about a topic can still produce polished pieces of work with AI, because it so easy to use. Such work can give the impression that students have learnt content when they haven’t – something that has been referred to as a mirage of false mastery.
AI use in schools is not limited to students. Three-quarters of teachers and school leaders have used AI for work. While student use raises concerns, teacher use of education-specific AI has strong potential to strengthen professional development and make tasks, such as administration, marking and feedback to students, more efficient.
By contrast, general-purpose AI does not exclusively draw from the evidence on how students learn best and can provide low-quality teaching and curriculum advice. Unfortunately, our data shows that ChatGPT (a general-purpose AI) is what is most commonly used by teachers and school leaders. This needs to change.
The most pressing problem is to ensure the integrity of senior secondary assessment and the awarding of qualifications and marks. Senior secondary assessment regularly includes both external assessment (often end-of-year exams) and internal, school-based assessment that occurs throughout the year.
These assessments are generally designed, administered and marked by individual schools. If students complete internal assessment in unsupervised settings where they can potentially access AI, it brings into doubt the accuracy of assessment and awarding of grades. None of this is easy for schools, and about half of all secondary teachers are concerned they don’t know how to prevent AI-related plagiarism or cheating.
Historically, a year 11-12 student’s best chance for getting into their preferred university or TAFE course has been to work hard to gain the marks needed for admission.
But now, what if they miss out on that course because the place went to a student who used AI to improve their assessments? Each Australian school system needs to urgently review assessment practice and awarding of marks and qualifications in senior secondary education. If post-secondary institutions, employers, the general public and students themselves lose confidence in senior secondary assessment systems, it will be very hard to win that confidence back.
AI is impacting all of secondary education (teachers report much lower use by students in primary school). Schools are changing how they assess students, moving away from assessments that are done at home and therefore more susceptible to AI use.
Slightly fewer than three-quarters of secondary teachers say they now assess students in class using pen and paper. Some schools are using other checks and balances, such as interviewing students about essays or projects done at home to gauge if they have completed the work themselves or outsourced it to AI.
Responding to AI means we need to change assessment practice. But we should not let our response to AI drive everything. We don’t want to get to a point where we say students should never do work at home; should never write extended essays or undertake large projects. Such tasks are essential for the breadth and depth of learning and improving crucial skills such as writing.
Too often, technology is put into the hands of children long before harmful effects are known and acknowledged. We can’t make the same mistake again. There is little evidence of the benefit of AI to student learning but clear research on the harm it can do. We must act now.
Image credit: János Szüdi