Generative AI is making it harder for teachers to know whether take-home work reflects a student’s own understanding, a new Australian report says. In a survey across 200 New South Wales schools, 61 per cent of high-achieving students said AI was eroding their competitive advantage, while 53 per cent of teachers supported temporarily suspending take-home assessments.
Watch Desk analysis
What happened
The Greek Herald reports that former NSW Education Standards Authority chair Tom Alegounarias sees AI as sharpening long-standing concerns about fairness in take-home work. Nearly two-thirds of students surveyed said their peers were using AI chatbots to complete homework or take-home assessments.
The research, conducted by assessment platform Gradeo, also found that 91 per cent of teachers believed AI-detection tools were no longer sufficient and assessment methods needed redesigning. Those are survey findings about respondents’ views, not a direct measure of how often students use AI or whether particular submissions were generated by it.
The report says some NSW schools have introduced more oral examinations, handwritten tasks and supervised assessments. NSW has also limited take-home Higher School Certificate assessments to one per course, worth no more than 7.5 per cent of a student’s final mark.
Why it matters
The pressure is not simply to catch students using chatbots. If work completed at home no longer gives teachers a reliable picture of what a student understands, assessment may need to shift towards supervised work, oral explanation or other ways of showing learning.
That shift has trade-offs. Take-home work was intended in part to ease pressure from formal exams, and changing assessment methods takes time and care. A detection tool that teachers say is no longer sufficient is not much of a policy plan on its own.
Our read
The useful signal here is the gap between the pace of AI adoption and the confidence teachers have in assessing work done outside the classroom. Restricting take-home marks and adding supervised or oral tasks are concrete responses, though the survey does not establish which approach produces the fairest results.
Schools should be able to explain what each assessment is meant to measure, and how students can demonstrate their learning when AI tools are readily available. “Just detect it” appears to be losing its appeal, and not a moment too soon.
What to watch
- Whether NSW schools adopt more supervised, handwritten or oral assessments.
- How states put broader restrictions on generative AI in classroom assessments into practice.
- Whether future research measures actual use of AI in student work, rather than confidence and reported perceptions.
Discussion spark: Should schools put more weight on supervised exams and oral assessment to preserve trust, even if that means less take-home work?
Sources and evidence
- Source update (10 November 2026, 04:31 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.