April, 2026 - TeacherTalk.site

From Efficient to Transformative: AI, SAMR and TPACK

Week 13 · Pedagogy · Teaching with Technology

From Efficient to Transformative: AI, SAMR and TPACK

“The question is not how can I use AI? — it is how can AI help me create a transformative, student-centred learning experience?“

⏱ 5 min read · By Matt Faulkner · Pedagogy / Teaching with Technology

Walk into any staffroom right now and AI is in the conversation. Some teachers are experimenting enthusiastically, others are cautious, and most are quietly trying to work out how to respond without losing what makes their teaching meaningful.

In a recent Faculty Focus article — Humanizing Generative AI: Three Ways to Keep Students at the Center of Your Classroom — Dr Richard Violanti names a pattern that should sit at the heart of every AI-in-education conversation we are having this year. Teachers are using generative AI to draft emails, build slides, summarise readings and generate quizzes. All useful. All versions of what we already did. As Violanti puts it, the technology “enhances the product … but leaves the process of learning for the students unchanged.”

That diagnosis deserves our attention. If our ambition stops at efficiency, we will have adopted the most powerful learning technology in a generation and used it to make a slightly faster worksheet.

If our ambition stops at efficiency, we will have adopted the most powerful learning technology in a generation and used it to make a slightly faster worksheet.

This post is a follow-up to Violanti’s piece. The aim here is to take his core argument — that AI’s real promise is pedagogical, not administrative — and sit it alongside two of the most widely used frameworks in educational technology: the SAMR model (Puentedura, 2006) and the TPACK framework (Koehler & Mishra, 2009). Together, they give us a sharper way to see when AI is genuinely transforming learning, and when it is quietly substituting for what was already there.

The efficiency trap, named

Violanti’s observation is not new, but it is important. Decades of educational technology research — from Larry Cuban onwards — show that most classroom innovations end up sustaining existing practice rather than transforming it. Interactive whiteboards replaced chalkboards. PDFs replaced photocopies. Learning management systems replaced filing cabinets. Each was a genuine workflow improvement. None, on its own, changed what learning felt like for students.

Generative AI is arriving with the same gravitational pull. The path of least resistance is to wire it into what we are already doing: a faster quiz generator, a tidier email reply, a cleaner slide. These uses are not bad — they give time back. But if that is where we stop, AI becomes another tool that makes the teacher’s life marginally easier without making the student’s learning meaningfully different.

The question Violanti poses is the right one, and it links directly to the kind of intentional planning we have talked about in earlier posts on planning with purpose: not how can I use AI?, but how can AI help me create a transformative, student-centred learning experience?

SAMR: a mirror for how we are actually using AI

Ruben Puentedura’s SAMR model offers four levels of technology integration: Substitution, Augmentation, Modification and Redefinition. The lower two levels enhance existing tasks. The upper two transform them — they make possible learning experiences that simply could not exist without the technology.

Mapped onto generative AI, the levels become uncomfortably clear.

🪜 SAMR applied to generative AI

Substitution — AI writes the quiz you used to write yourself. Same quiz, same pedagogy, faster delivery.
Augmentation — AI writes a better quiz: adaptive difficulty, auto-generated distractors, instant feedback. Still a quiz.
Modification — A reading becomes a conversation. Students use an AI tutor to interrogate a text, surface their own misconceptions, and bring questions to class. The task is genuinely redesigned.
Redefinition — A task that could not exist before. A history student conducts a simulated interview with a historical figure and has to defend, with evidence, where the AI’s responses sit in the record and where they drift into plausible-sounding fiction.

None of this requires teachers to become SAMR theorists. The value of the model is its reflective prompt, which Violanti phrases neatly: is technology making learning different, or just more efficient? If the answer is “just more efficient,” we are at the bottom of the ladder. That is a reasonable place to start. It is not a reasonable place to stay.

It is worth being honest about SAMR’s limits too. Critics have pointed out, fairly, that the model can privilege product over process and can imply Redefinition is always pedagogically superior, which it is not. A well-designed Substitution task can teach more than a poorly-designed Redefinition one. SAMR is a mirror, not a scoreboard.

TPACK: why the mirror alone is not enough

If SAMR asks where on the ladder are we?, the TPACK framework asks what do we need to know to climb it well? Koehler and Mishra’s model describes teaching as the intersection of three knowledge domains: Content Knowledge (what we teach), Pedagogical Knowledge (how we teach) and Technological Knowledge (the tools we use). Transformative teaching lives at the centre of the Venn diagram, where all three overlap.

This matters enormously for AI, because AI is the first mainstream classroom technology where the Content and Pedagogy circles are genuinely under pressure from the Technology circle. A generative model can plausibly do some of what we used to teach students to do — write an essay, summarise an argument, solve a routine problem. TPACK reminds us the answer is not to retreat into one circle or another. It is to redesign the intersection.

🎯 Three questions to ask — in this order

Content — What is the deep disciplinary knowledge I want students to build, and what parts of that knowledge are now more, not less, important because AI can approximate the surface of it?
Pedagogy — What does student-centred learning of that content actually look like — inquiry, dialogue, critique, production, reflection?
Technology — Where, specifically, does AI deepen that pedagogy rather than shortcut it?

When those three questions are asked in that order, AI stops being the starting point and becomes a design choice. That is TPACK doing its quiet, useful work.

What transformation actually looks like

Violanti offers three concrete practices that, in SAMR terms, push teaching toward Modification and Redefinition, and in TPACK terms sit firmly inside the intersection of all three circles. They are worth restating — and building on.

1. Differentiate, don’t just generate. Instead of producing one more artefact for the whole class, use AI to produce multiple on-ramps to the same content: a levelled version of a dense reading, an audio summary, a concept map, a worked example with scaffolding, a stretch version with primary sources. Every student engages with the same learning goal, but through the modality and entry point that works for them. This is differentiation at a scale that was genuinely impossible a few years ago, and it sits squarely in Redefinition territory.

2. Turn AI into a thinking partner, not an answer machine. The concern that AI short-circuits thinking is legitimate, but it is a design problem, not a technology problem. Assignments that ask students to produce with AI tend to flatten thinking. Assignments that ask students to interrogate AI tend to deepen it. Have students debate an AI-generated argument, diagnose where a model oversimplified a concept, compare AI output against primary sources, or co-develop a solution and then annotate which contributions were theirs and which were the model’s. This connects directly to the kind of higher-order questioning we have explored in earlier posts — the critical move is requiring students to evaluate AI, not just consume it. Metacognition becomes the assessable product.

3. Use AI to design, not just to deliver. This is the move most teachers have not yet made. Ask AI to help you generate learning experiences at different cognitive levels, propose authentic problem scenarios linked to your subject, suggest formative assessments aligned to specific outcomes, or role-play a sceptical student and stress-test your explanations. The teacher stays the designer. The AI becomes a collaborator in the design process itself. This is perhaps the highest-leverage shift available right now — and the one most hidden from view, because it happens before the lesson even starts. It pairs particularly well with what we have written previously about formative assessment: AI can help generate the next layer of probing questions after you see how students responded to the first.

A few honest caveats

A responsible follow-up has to acknowledge what this kind of transformation does not solve. Designing transformative AI tasks takes more time up front, not less — the efficiency trap is seductive partly because it is genuinely easier. It also assumes equitable student access to tools, clear school guidance on academic integrity, and professional learning that treats AI as a pedagogical question rather than a software training session. None of that is evenly distributed. And the technology itself is not neutral: generative models carry bias, hallucinate confidently, and raise real questions about labour, intellectual property and environmental cost that belong in any serious classroom conversation about their use.

Holding those caveats in mind is not a reason to retreat to Substitution. It is a reason to make the climb to Redefinition deliberately, transparently, and with students as partners in the conversation.

The more human use of artificial intelligence

The irony of generative AI in education is that its most powerful contribution may be the thing it cannot do. It cannot know a student. It cannot notice the one who has gone quiet in week six. It cannot decide what in a subject is worth knowing. It cannot replace the trust that makes feedback land.

What it can do — if we let SAMR and TPACK do their reflective work — is give that time back to us. Time to design richer tasks. Time to respond to individual students. Time to do the pedagogical work that most drew us to teaching in the first place.

Teaching has never been about the tools. It is about the people in the room. The measure of whether we are using AI well is not how much time it saves us — it is whether the learning that happens in that room is different, deeper, and more clearly centred on the student than it was before.

If it is, we have climbed the ladder. If it is not, we have simply bought a faster worksheet.

Your turn.Where are you on the SAMR ladder with AI right now — still mostly Substitution, or pushing into Modification and Redefinition? Drop a comment below with one AI-supported task you have redesigned (or want to redesign) this term. I would love to hear what is actually working in real classrooms.

References and further reading

Koehler, M. J., & Mishra, P. (2009). What is technological pedagogical content knowledge (TPACK)? Contemporary Issues in Technology and Teacher Education, 9(1), 60–70.

Puentedura, R. R. (2006). Transformation, technology, and education. Hippasus.

Hamilton, E. R., Rosenberg, J. M., & Akcaoglu, M. (2016). The Substitution Augmentation Modification Redefinition (SAMR) model: A critical review and suggestions for its use. TechTrends, 60(5), 433–441.

Violanti, R. J. (2026). Humanizing generative AI: Three ways to keep students at the center of your classroom. Faculty Focus.

AI and the Future of Learning: What We Keep, What We Redesign

In Part 1 we grappled with what AI is and the ethical questions it raises. In Part 2 we looked honestly at what’s happening in schools right now. Now comes the hardest part: imagining forward.

If we had the opportunity to redesign education with AI as a given — not a disruption to manage, but a reality to build around — what would we keep? What would we deliberately redesign? And what would we fight, with everything we have, to protect? These are not abstract questions. The decisions being made in schools, systems, and staffrooms right now are shaping the answer.

Harnessing a Technology That Won’t Wait

One of the most disorienting things about generative AI is its pace of change. The tools available to teachers and students today are significantly more capable than those available eighteen months ago, and the tools available eighteen months from now will likely make today’s seem primitive. Education systems are built for decade-long policy cycles. AI is operating on a month-to-month development timeline.

This is not a reason for paralysis. It is an argument for building adaptive capacity — in teachers, in schools, and in systems — rather than chasing specific tool competencies that will be outdated before the professional development program is complete. What teachers need is not training in ChatGPT version 4. What they need is a deep enough understanding of how AI works, what it can and cannot do, and what good learning looks like, to make intelligent, principled decisions as the technology continues to evolve.

Australia’s senior educational policymakers have identified managing risks, educating teachers, and educating system leaders as the highest priorities for responsible AI integration. The research is clear that teacher capacity is both a top priority and a major challenge — and that ways for teachers to undertake significant professional learning to equip themselves and their students with AI skills, without overburdening them, remain genuinely unclear. This is the honest state of the field.

“We do not need teachers who can use every AI tool. We need teachers who understand learning deeply enough to know when AI helps it and when it harms it.” — TeacherTalk

teacher and students in a future-facing, collaborative classroom setting — The classroom of the future will be defined less by technology and more by the quality of the human relationships and thinking within it

What Needs to Be Redesigned: Assessment

If there is one part of the education system that AI has most urgently exposed as in need of redesign, it is assessment. The traditional essay, the take-home assignment, the multiple-choice test — these formats were designed for a world in which students had to do their own thinking to produce a response. In a world where AI can generate a credible essay on any topic in thirty seconds, assessing a student’s learning through a written product alone is no longer adequate.

This doesn’t mean abandoning writing. Writing is thinking made visible, and the process of composing a careful argument remains one of the most powerful learning activities a student can undertake. But it does mean reconsidering what we assess, and how we authenticate it.

Assessment reform for the age of AI is pointing in several consistent directions. Performance-based and oral assessments — where students must demonstrate understanding in real time, in conversation, with follow-up questions — are inherently AI-resistant and arguably more valid measures of genuine learning anyway. Portfolio assessment that captures process, including drafts, reflections, and revision histories, provides evidence of thinking that AI cannot fabricate. Authentic tasks with local, specific, personal dimensions — the kind of performance of worth we’ve written about before — demand contextual knowledge and personal voice that generic AI outputs cannot replicate.

These approaches are not new. They are, in many cases, what good assessment has always looked like. AI has simply made the case for them more urgent.

✅ Assessment Design Principles for an AI World

Build in process evidence. Draft submissions, reflective journals, and annotated bibliographies document thinking that AI cannot fake.
Use local and specific context. Tasks anchored to the student’s own community, experience, or school are far more AI-resistant than generic topics.
Include oral or performance components. A student who can explain their work, respond to follow-up questions, and articulate their reasoning has demonstrably done the thinking.
Separate the product from the learning. What did the student learn, not just what did they produce? Exit tickets, verbal check-ins, and the diagnostic questioning strategies in last week’s post on reading the room are more relevant than ever.
Design with the “so what” in mind. Assessments connected to real-world purpose, real audiences, and genuine consequences for the community are inherently more meaningful — and more AI-resistant. Revisit our post on the “so what” of learning.

What AI Could Do for Personalised Learning — If We Get It Right

Here is the genuinely exciting possibility. For decades, educators have understood that different students learn differently, at different rates, with different entry points and different strengths. The practical constraint has always been scale — one teacher, thirty students, one lesson. Differentiation has been the aspiration, but genuine personalisation at scale has been largely impossible.

AI changes this constraint. Adaptive learning platforms that adjust in real time to student interactions can, in principle, provide every student with an experience calibrated to their current understanding — extending the student who has mastered the concept, supporting the student who needs a different explanation, and providing the student with a disability or language barrier with accommodations that would otherwise require specialist support.

This is not science fiction. Carnegie Learning’s adaptive mathematics platform, Khan Academy’s Khanmigo tutoring tool, and government-built tools like NSW’s EduChat are all operational examples of this direction. Carnegie Learning’s adaptive learning platform provides a customised learning experience that aims to adapt in real time to each student’s interactions, while Khan Academy’s Khanmigo offers AI one-on-one tutoring by mimicking a writing coach — giving prompts and suggestions to move students forward as they write, debate, and collaborate.

The equity caveat, however, remains essential. Access to these tools is not uniform. Students in well-resourced urban schools with reliable devices and fast internet are positioned to benefit most from AI personalisation. Students in under-resourced schools, remote communities, and low-income families may find that the digital divide that has always characterised Australian education simply takes a new form. Without deliberate intervention — in infrastructure, in teacher training, in culturally responsive design — AI personalisation risks becoming another advantage for the already advantaged.

The Teacher in an AI World: More Important, Not Less

Perhaps the most important thing to say about the future of education and AI is this: the teacher becomes more important, not less, as AI becomes more capable.

This seems counterintuitive. If AI can deliver content, provide feedback, personalise the learning pathway, and assess the outcome — what is left for the teacher? The answer is everything that matters most.

The teacher who knows which student is being bullied and is too ashamed to ask for help. The teacher who notices that a child’s sudden academic disengagement reflects something happening at home. The teacher who creates the conditions of psychological safety that allow a student to take an intellectual risk. The teacher who connects a student’s passion for gaming to a career pathway they had never imagined. The teacher who, through the quality of their relationship with a child, communicates the fundamental message: you are capable, you matter, and I believe in you.

No AI does any of this. And in a world where AI handles more of the cognitive transactional work of education, these deeply human capabilities become the irreplaceable core of what teachers are for. The profession must articulate this loudly and clearly — not as a defensive reaction to AI, but as a confident statement of what teaching actually is.

teacher in genuine one-on-one conversation with a student, warm and engaged — No AI replicates the teacher who knows a student’s story and believes in their potential. That is the irreplaceable core of teaching.

The Questions That Will Define the Next Decade

We began this series with a set of hard questions. It is worth returning to them, not with definitive answers — the field is moving too fast for those — but with the honest acknowledgement that these questions demand deliberate, ongoing engagement from every educator.

🔍 The Questions That Will Shape Education’s AI Future

Can AI help teachers reach more kids and create better learning experiences? Yes — if we design the tools with equity at the centre, invest in teacher capability, and use AI to amplify human teaching rather than replace it.
Will AI hurt or help the traditional model of schooling? Both. It will expose what is worth keeping and accelerate the demise of what was never good enough. That is, ultimately, an opportunity.
Who is monitoring and researching the impact of AI on students, teachers, and the system? Researchers, governments, and frameworks are emerging — but the evidence base is thin and the technology is moving faster than the research. Schools must become active generators of evidence, not passive recipients of it.
What makes this different from every other tech revolution in education? AI operates inside the learning process itself. It doesn’t just present information — it can do the thinking. That is qualitatively different from every previous technology, and demands a qualitatively different response.
How can we harness a technology that is evolving so exponentially? By investing in deep pedagogical knowledge and adaptive professional capacity, not tool-specific training. Teachers who understand learning deeply will make good decisions about AI, whatever form it takes.
What happens when AI designs, completes, and grades the task? Nothing is learned. The answer is not to ban AI, but to redesign tasks so that the human thinking is visible, valued, and authenticated — and to build a school culture where genuine learning is the goal, not just the credential.

A Final Word: What We Fight to Protect

Education is not primarily about information transfer. It never was. It is about the development of human beings — their capacity to think, to connect, to create, to care, and to contribute to a world that is shared with others. AI can support some of those processes. It cannot replace any of them.

What we fight to protect, in an AI-saturated world, is the space for genuine human thinking — the productive struggle, the creative leap, the moment of understanding that belongs to the student and to no one else. We fight for equitable access to the benefits AI offers, because a technology that advantages the already advantaged is not progress. We fight for First Nations students whose cultures, languages, and ways of knowing must not be erased or distorted by AI systems trained on predominantly Western datasets. And we fight for the teaching profession — not as it is, but as it could be at its best: relationships of trust, built over time, between adults who believe in children and children who are learning to believe in themselves.

That is what education is for. AI does not change it. It simply asks us to be clearer about it than we have ever had to be before.

For practical strategies to keep human thinking at the centre of your lessons, revisit our posts on higher-order questioning, formative assessment, and assessment for, as, and of learning.

“AI will not replace great teachers. But teachers who understand AI will replace those who don’t — because their students will be better prepared for the world that actually exists.” — TeacherTalk

💬 Your Turn! What would you keep, redesign, or fight to protect in education as AI reshapes the landscape? This is one of the most important conversations our profession has ever had — and your voice belongs in it. Share your thoughts below.

References

Bower, M., Hedberg, J., Kuswara, A., & Dalgarno, B. (2025). What generative artificial intelligence priorities and challenges do senior Australian educational policy makers identify (and why)? The Australian Educational Researcher. https://doi.org/10.1007/s13384-025-00801-z

Furze, L. (2025, December 3). What the National AI Plan means for schools. leonfurze.com. https://leonfurze.com/2025/12/03/what-the-national-ai-plan-means-for-schools/

Holmes, W., Porayska-Pomsta, K., Holstein, K., Sutherland, E., Baker, T., Shum, S. B., Santos, O. C., Rodrigo, M. T., Cukurova, M., Bittencourt, I. I., & Koedinger, K. R. (2023). Ethics of AI in education: Towards a community-wide framework. International Journal of Artificial Intelligence in Education, 32, 504–526. https://doi.org/10.1007/s40593-021-00239-1

Lodge, J. M., Bower, M., Gulson, K., Henderson, M., Slade, C., & Southgate, E. (2025). Australian Framework for Artificial Intelligence in Higher Education. Australian Centre for Student Equity and Success. https://www.acses.edu.au

Lodge, J. M., Howard, S., Selwyn, N., & Corrin, L. (2023). Assessment reform for the age of artificial intelligence. TEQSA. https://www.teqsa.gov.au

National AI in Schools Taskforce. (2023, updated 2024). Australian Framework for Generative AI in Schools. Australian Government Department of Education. https://www.education.gov.au

Varsik, S., & Vosberg, L. (2024). The potential impact of artificial intelligence on equity and inclusion in education. OECD Artificial Intelligence Papers, No. 23. https://doi.org/10.1787/15df715b-en

AI in Classrooms Right Now: The Good, the Messy, the Urgent

In Part 1 of this series, we asked the hard questions about what AI is and what ethical challenges it poses for Australian schools. Now let’s get specific. Because AI in education is no longer theoretical — it is happening in classrooms right now, in ways that are extraordinary, messy, promising, and at times deeply confronting.

This post looks at what AI is actually doing in schools today: the genuine benefits being documented in practice, the scenarios that are challenging everything we thought we knew about assessment and academic integrity, and the question that sits at the centre of it all — when AI can design the task, complete the task, and grade the task, what exactly is the student doing?

What AI Is Actually Doing in Australian Schools

Australia is moving faster on AI in schools than many realise. Research conducted in 2024 found that 78.2% of secondary schools were already using AI in some form. The nature of that use varies enormously — from a teacher using AI to generate a first draft of a rubric, to a school deploying a government-built AI tutoring assistant for every student.

The most significant state-level examples are worth knowing about:

🏫 AI in Australian Schools: Current Examples

NSW — EduChat: Trialled in 16 schools in Term 1 2024, expanded to 50 schools in Term 2, rolled out to all NSW public school staff by the end of 2024, and now available to all Year 5–12 students. Hosted in Azure’s Sydney data centre, it uses a blend of OpenAI models and eight content filtering systems. Crucially, it is designed to refuse direct answers — instead responding with guiding questions to encourage critical thinking. The NSW Department explicitly states that free-to-use commercial AI tools are neither endorsed nor recommended for use in their schools.
South Australia — EdChat: Rolled out to all SA public high schools from October 2025. Functions as an AI tutoring assistant helping students explore their coursework more deeply, with government oversight of data and content.
Queensland: The Queensland Department of Education has emphasised generative AI for regional and remote student populations, using AI to enable personalised learning, differentiated support, and improved accessibility in areas where teacher resources are limited.
Khan Academy — Khanmigo: Used in some Australian schools, this AI tutoring tool offers one-on-one coaching by mimicking a writing coach — giving prompts and suggestions to move students forward as they write, debate, and collaborate, rather than simply providing answers.
Carnegie Learning: An adaptive learning platform used in mathematics education that adapts in real time to each student’s interactions, providing a customised learning experience responsive to varied learning requirements.

Can AI Help Teachers Reach More Kids?

This is the question that genuinely excites many educators — and the evidence, while still early, is encouraging in specific areas.

AI tutoring tools can provide a form of one-on-one support that no teacher can replicate at scale. In a class of 28 students, a teacher can interact meaningfully with each child for perhaps a few minutes per lesson. An AI tutoring tool can engage every student simultaneously, adapt to their current understanding in real time, and provide immediate, individualised feedback. For students with English as an additional language, for students with learning difficulties, for students in remote schools with limited specialist support, these capabilities represent a genuine opportunity to reduce disadvantage.

AI can also free up teacher time for the work that only humans can do. When AI handles routine tasks — generating first drafts of lesson materials, creating differentiated worksheet versions, producing formative quiz questions, or providing initial written feedback on drafts — teachers can redirect their attention to relationship-building, targeted small-group instruction, and the nuanced professional judgement that is at the heart of great teaching. Our earlier posts on formative assessment and designing a performance of worth are directly relevant here — AI can support these processes, but it cannot replace the teacher’s judgement in designing tasks that are worth students’ genuine effort.

Most high school teachers who used AI in their work in the 2024–25 school year reported that these tools saved time and improved the quality of their work. That finding matters — teacher workload is one of the most significant drivers of attrition from the profession, and anything that genuinely reduces administrative burden without compromising student outcomes is worth taking seriously.

The Scenario Nobody Wants to Talk About

Here it is. The scenario that is already happening in classrooms around the world, including in Australia. A teacher uses AI to design an assessment task. A student uses AI to complete that task. The teacher uses AI to provide feedback on the submission. And somewhere in that loop, the student has done very little actual thinking.

This is not a hypothetical. Research examining the impact of generative AI on assessment has found that assessments were easily compromised using generative AI, and that many markers found it challenging to identify which student work was human-authored, which was AI-modified, and which was AI-authored. In higher grade bands, students using AI can generate work that is not only undetectable but that achieves grades their own unaided writing would not have reached. For teachers, this creates an unsettling dynamic where the marking process becomes contaminated by uncertainty — second-guessing authorship, with grade disparities emerging between markers who suspect AI use and those who do not.

In one telling study at a UK business school, markers showed unusually large grade disparities between first and second markers — with some appearing to have consciously or subconsciously marked down work they suspected was AI-generated. This is not a defensible assessment practice, but it reflects a very human response to a genuine problem of attribution in an AI-saturated environment.

“When AI designs a task, the student uses AI to complete the task, and AI grades the task for the teacher — what exactly has been learned? This is not a future scenario. It is happening now.” — TeacherTalk

The Cognitive Offloading Problem

There is a subtler risk beneath the academic integrity conversation that deserves more attention: cognitive offloading. Research is finding that students who rely on AI to complete higher-order thinking tasks may produce stronger immediate answers but develop weaker long-term learning and less independent capability. When novices lean on AI they often skip the hard metacognitive work — the struggle, the confusion, the revision — that builds genuine expertise.

A national survey of over 1,200 teens and young adults found that 41% believe AI will have both positive and negative impacts on their lives in the next decade. Those expecting negative impacts worry primarily about job displacement and AI dominance — but the educational risk of cognitive offloading is arguably more immediate. If students are using AI to avoid the hard thinking that builds neural pathways and conceptual understanding, the short-term grade gain may come at a long-term developmental cost.

This is precisely why the design of NSW’s EduChat is significant. By refusing to provide direct answers and responding instead with guiding questions, it attempts to position AI as a Socratic tutor rather than a task-completion engine. Whether that design holds up when students are stressed, time-poor, and looking for shortcuts remains to be seen.

What About Academic Integrity? The Research Is Mixed

The academic integrity question is genuinely complicated. Here is what the research actually shows:

Surveys of secondary students have found that approximately 19% of teens who are aware of ChatGPT have used it to help with schoolwork. The same research found that students consider using AI to research new topics broadly acceptable (69%), but consider using it to write an essay far less so (20%). There appears to be a student-held intuition about where AI assistance crosses a line — though that intuition is inconsistently applied under pressure to perform.

Interestingly, a study based on confidential student surveys found that cheating rates did not significantly increase in the early period of AI adoption — though a majority of teachers believe AI will negatively impact academic integrity over the next three years. The gap between what students are doing and what teachers fear they are doing is itself an important data point.

What is clear is that the solution is not solely technological. Detection tools using AI to detect AI-generated text are in an arms race with the generative tools themselves and will never be definitively reliable. The answer, as multiple researchers have concluded, is partly cultural — building a genuine school ethos around academic integrity, helping students understand the value of their own thinking, and redesigning assessments so that AI completion provides no meaningful advantage.

That last point brings us to the most important pedagogical response available to teachers right now: task design. Our post on the “so what” of learning is directly relevant here — tasks with a genuine real-world purpose, requiring personal voice, local knowledge, and specific evidence of process, are inherently more AI-resistant than generic essay prompts. The performance of worth is making a comeback.

Will AI Hurt or Help the Traditional Model of Schooling?

Both. The traditional model of schooling — teacher-delivered content, individual written assessment, grades as the primary currency of learning — is genuinely vulnerable to AI disruption. If the goal of education is to produce a correct written response, AI can do that faster and often better than a student can. That should prompt a serious rethink of what we’re assessing and why, not a defensive reaction against the technology.

But the elements of schooling that AI cannot replicate — human relationships, mentorship, the socialisation of learning, the development of character, curiosity, and resilience — become more important, not less, in an AI-saturated world. The teacher who knows a student’s story, who notices when something is wrong, who believes in a child’s potential before the child believes it themselves — no algorithm does that. And that is where the teaching profession should be investing its energy and its advocacy.

In Part 3 of this series, we’ll look forward: what could an AI-informed education system look like if we designed it well? What would we keep, what would we redesign, and what would we fight to protect? Stay tuned.

💬 Your Turn! Have you seen AI used well in a classroom — by a student or a teacher? Or have you encountered the scenario where AI did the work and nobody learned anything? Share your experience below — the most important research on this topic is happening in classrooms like yours.

References

Digital Education Council. (2024). Global AI student survey 2024. Digital Education Council.

Furze, L. (2025, December 3). What the National AI Plan means for schools. leonfurze.com. https://leonfurze.com/2025/12/03/what-the-national-ai-plan-means-for-schools/

Kofinas, A., Malesios, C., Vikan, A., & Lev-On, A. (2025). The impact of generative AI on academic integrity of authentic assessments within a higher education context. British Journal of Educational Technology. https://doi.org/10.1111/bjet.13585

Pew Research Center. (2023). Teens and AI: Awareness and use of ChatGPT. Pew Research Center.

Teaching for Tomorrow. (2025). AI in high school education: Trends, challenges, and opportunities. Bowdoin College Hastings AI Initiative.