A professor specialising in the uses and misuses of AI in education said new software to detect it in students\u2019 work could not be trusted, and teachers might need to rethink how they assess pupils\u2019 learning entirely.<\/p>\n
The Principals\u2019 Federation is calling on the Government to provide a firmer national strategy, with claims that teachers are \u201cthirsting for knowledge\u201d on how to tackle the perpetually changing landscape.<\/p>\n
Meanwhile, the Education Review Office and the Ministry of Education have launched research to guide schools as they try to strike a balance between teaching students to use AI safely and equipping them for the future.<\/p>\n
Principals share their AI tactics<\/p>\n
So how are our schools coping with the rise of AI?<\/p>\n
Wakatipu High School principal Oded Nathan said the school was well prepared for dealing with AI, but his biggest fear was students failing to develop critical thinking skills.<\/p>\n
\u201cWe\u2019re super cognisant of the changing world. AI is not going to disappear; it\u2019s not going away.<\/p>\n
\u201cOur kids are using it, our community\u2019s using it, our business leaders are using it.\u201d<\/p>\n
The school had introduced an AI traffic light system \u2013 green allowed full use, orange limited use, as outlined in the task instructions, and red banned it. <\/p>\n
Teachers had completed two days of training on the framework and how to coach students to use AI healthily. The framework was not about erasing AI, but ensuring it enhanced, rather than replaced, students\u2019 work.<\/p>\n
ChatGPT had been removed from the school\u2019s Wi-Fi. As a Google school, it allowed access to Gemini, which had guardrails to stop students from completing entire assessments with AI. <\/p>\n
Some schools are banning ChatGPT from the school networks to try to combat the rise of AI. Photo \/ 123RF<\/p>\n
At Papakura High School, principal Simon Craggs said a similar traffic light system had been adopted.<\/p>\n
\u201cAI is a tool that is already revolutionising the workplace, and we cannot afford to leave students unprepared for this new reality.\u201d<\/p>\n
Staff had received training to build confidence with AI tools. <\/p>\n
\u201cAcademic misconduct has not yet been a major problem for us and has been relatively easy to pick up because of the teachers\u2019 knowledge of the students.\u201d<\/p>\n
At Whang\u0101rei Girls\u2019 High School, principal Sonya Locker said AI misuse in lower-stakes assessments had been treated as a \u201cteachable moment\u201d. More serious breaches had been investigated by senior leaders.<\/p>\n
\u201cAt times, students can give evidence of the authenticity of their work through notes and workings. If the investigation concludes that a breach has occurred, the student is reported to NZQA.\u201d<\/p>\n
\u014ct\u016bmoetai College principal Russell Gordon said the goal had been to spark a wider conversation about digital literacy and academic responsibility.<\/p>\n
\u201cThe goal is not to create fear around new technology, but to help students understand how to use emerging tools ethically and appropriately within defined conditions.<\/p>\n
\u201cThe focus should remain on authentic learning, equitable assessment, and preparing young people to engage responsibly with the technologies shaping their future.\u201d<\/p>\n
Software to check for AI is \u2018problematic\u2019<\/p>\n
Jason Stephens, Auckland University professor of psychological studies in education, who has studied the uses and misuses of AI, said he did not trust \u201cproblematic\u201d software designed to check students\u2019 work.<\/p>\n
\u201cI\u2019ve heard anecdotal reports of people submitting their own work that they wrote pre-AI and it getting flagged. And it\u2019s really just too easy to manipulate.\u201d<\/p>\n
A smart student would run AI-produced work through a programme to make it sound more like their own, he said.<\/p>\n
\u201cWhether that\u2019s to run it through a humaniser or make some changes themselves to make sure that there\u2019s a little bit of a human touch in there to throw it off.\u201d<\/p>\n
Stephens said the rise of generative AI had transformed the way teachers assessed students\u2019 work. <\/p>\n
Fears of an AI echo chamber <\/p>\n
AI lead at Deloitte, Dr Amanda Williamson, said it was important to teach students about AI as soon as they began interacting with it.<\/p>\n
\u201cFor young people and the adults around them, sit with them. Show them what AI actually is and what it is not. Help them build instincts.\u201d<\/p>\n
Williamson highlighted a major danger for students using AI: its sycophantic nature.<\/p>\n
She said the technology was built to agree with you.<\/p>\n
Deloitte artificial intelligence director Dr Amanda Williamson<\/p>\n
\u201cIt will validate a wrong answer, reinforce a bad assumption and, by default, tell you your work is excellent when it is not.<\/p>\n
\u201cAn adult with professional experience might catch that. A young person building their understanding of the world may not. Show them that. Let them see it happen.<\/p>\n
\u201cThat single lesson teaches more about AI than any policy document. That same instinct, knowing when to trust AI output and when to challenge it, is what the workforce needs.\u201d<\/p>\n
Principals\u2019 Federation calls for a strategy <\/p>\n
New Zealand Principals\u2019 Federation president Jason Miles wanted a firmer strategy on how to use AI in classrooms.<\/p>\n
\u201cWe need direction from the Government. We need professional learning from the Government about the equity issues around AI, the ethical and privacy issues around using AI, and how we can use it effectively to support students in their teaching and learning.\u201d<\/p>\n
He said there was an unfair expectation on teachers and principals to keep up with the changing landscape, including how to detect AI in students\u2019 work, how to use it to create lesson plans and how to incorporate it in student learning.<\/p>\n
What does the guidance say?<\/p>\n
Rob Mill, the Ministry of Education\u2019s group manager for curriculum, said the department had put out guidance on AI use.<\/p>\n
The guidance, first issued in 2024, highlighted the potential for AI to enhance teaching and learning, but warned that its use should be restricted.<\/p>\n
It also noted potential cultural biases, saying AI was built on \u201cdominant cultures and languages\u201d.<\/p>\n
\u201cThe tools may not accurately reflect indigenous knowledge. From a New Zealand context, they are likely to be weak on m\u0101tauranga and te reo M\u0101ori, as well as Pasifika languages and Polynesian cultures.\u201d<\/p>\n
The ministry was working proactively with schools to gather insights on the ever-changing AI landscape.<\/p>\n
Ruth Shinoda, the Education Review Office\u2019s acting chief executive and chief review officer, said it was researching how students, teachers and leaders used AI.<\/p>\n
\u201cIt will also look at the challenges associated with AI use and how schools are responding to these challenges.\u201d<\/p>\n
She said the results were expected to be published later this year.<\/p>\n