Artificial intelligence has moved into New Zealand classrooms faster than the rules meant to govern it, according to a report released on Thursday by the Education Review Office, which found more than nine in ten school leaders and more than four in five teachers are now using the technology while only about a quarter of schools have any rules telling students what they can and cannot do with it.
The report, titled Ready or Not, is the first national picture of how schools are handling generative AI. The review office collected its data in Term 1 this year, drawing on almost 4000 survey responses from students, teachers, school leaders and parents, along with interviews and focus groups involving 129 people and visits to eight schools. It covers English-medium state and state-integrated schools with students from Year 5 up.
The headline finding is how ordinary the technology has become. Ninety-three percent of school leaders reported using AI, most often for drafting communications, summarising information and creating learning resources. Around half use it to write school policies. Teachers use it mainly to tailor teaching material, with 79 percent doing so, and to plan lessons, at 65 percent. Far fewer, 22 percent, use it for marking and feedback, with teachers uneasy about a machine judging student work. Three in four students said they used AI for schoolwork, most commonly to find information and generate ideas.
Education Review Office chief executive David Ferguson told RNZ he was surprised by how far it had spread. “It’s been one of the biggest game changers in the sector I think in the last few years… We’re led to believe it changes quickly so we think it’s really significant,” he said.
What worries teachers is not the workload help but what happens to students who lean on it. More than half of teachers told the review office that AI use had already had a negative effect on students’ critical thinking. The concern was sharpest in secondary schools, where 71 percent said so, compared with 41 percent in primary schools. More than half also thought creative thinking was suffering, although almost half believed independent learning had improved.
Students said much the same thing about each other. “It gives you the answer, but you don’t do any thinking for yourself,” one told the review office. Another described it as a good tool that some of their classmates lean on too heavily. “I use it to get inspiration for when I’m doing my work. But at the same time, I know students who use it as a crutch to do all their work for them,” they said. A primary school pupil put it more bluntly, saying it was addictive and that once you used it once you could not stop.
The report is careful to say the outcome depends on how the tool is used rather than whether it is used at all. Teachers and students who described using AI to check work, solve problems or spark ideas were more likely to report positive effects on thinking. The negative reports clustered around students who used it to produce the work itself.
Misuse is common enough to be a live problem for secondary teachers. Thirty-eight percent of secondary students admitted using AI for schoolwork in ways they knew they should not, alongside 19 percent of primary students. In assessments the figures were lower, at 19 percent of secondary students and 11 percent of primary. Three in four secondary teachers said they had come across inappropriate student use at least occasionally. Teachers who had struck it were five and a half times more likely to say AI had increased the time they spent checking whether work was genuinely a student’s own.
The review office also found a fairness problem building underneath all this. Students with stronger literacy and digital skills are using AI in more sophisticated ways that teachers find harder to detect, which the report says is creating an equity gap in who gets away with it. Ferguson said the review did not find the digital divide he expected between schools of different socio-economic backgrounds. “We didn’t discern differences there in terms of equity or a digital divide. It was so widespread we thought, that it looks like it’s being used far and wide,” he said.
Where the report is most critical is on the absence of rules. Around three in four schools were classed as taking an unplanned approach rather than a strategic one. Just 28 percent of leaders said they had policy or guidance for teacher use of AI in teaching. Only 26 percent had rules for student use in learning, with a quarter of leaders saying individual teachers were left to set their own expectations. Around seven in ten leaders said they had no policy covering data privacy and security when teachers use these tools, a gap that matters given how much student information passes through them.
Assessment is the exception, because the New Zealand Qualifications Authority already sets expectations for NCEA. Seventy-seven percent of secondary school leaders said they had an assessment policy in place and the rest said they were writing one. Fewer than one in four primary leaders had anything comparable. Having a policy is not the same as having a useful one, though. Of secondary students whose schools had assessment rules, only 43 percent thought those rules actually helped them work out what was allowed.
Teachers said they were flying blind themselves. More than half of leaders and 45 percent of teachers said they did not know enough to use the tools properly, a pattern the report found across schools of every size, decile and location, which it called a system-wide capability gap. About half of leaders and teachers said they lacked clear guidance. Roughly half of both groups said the tools were not yet trustworthy or reliable. One person offered a memorable summary of that unreliability, saying you can search up something about a butterfly and it will tell you about pickles while showing you a picture of a strawberry.
A small group has opted out altogether. Fifteen percent of teachers do not use AI in their work, and the report describes some of them as conscientious objectors who refuse on principle, worried about the example it sets and its effect on creativity and independence.
Parents are not asking for it to be banned. Eighty-four percent said they wanted their children taught how AI actually works, and 75 percent wanted them taught how to use it properly for schoolwork. Half of students said the same. Yet only 31 percent of leaders said their school explicitly teaches students about AI at all.
The review office has made nine recommendations, the central one being a national AI in education framework drawn from international models but adapted for New Zealand, along with consistent national policies for teacher and student use, guidance for parents, professional development for teachers, and an investigation into whether the government should provide approved AI tools to schools centrally. It recommends weaving AI literacy through existing subjects rather than creating a standalone AI subject for senior students, an option the government was weighing last year.
Ferguson said a common approach would help. “This is a big deal for schools in terms of keeping everybody safe, making sure they’re maximising the potential in terms of using it. That’s why we’re suggesting that we think having a consistent deliberate, strategic framework for all schools, a centrally-provided one, would be a good idea,” he said.
One teacher quoted in the report captured the balance most schools are trying to strike. “Something that can lower the workload is always welcome. We just have to make sure that we’re doing it safely. Teaching needs a teacher, and you simply can’t replace that.”
Has AI changed the way your children do their homework, or the way you teach? Tell us what you are seeing in your school in the comments below.
This article was written by AI, briefed to report the facts, hopefully without some of the bias people bring to the job π
Leave a Reply