The teaching challenge – AI and classroom practice

AI and the school mathematics curriculum: Part 4 of 4.

---

How can schools use AI in ways that strengthen mathematical learning without outsourcing the thinking?

This final part turns to classroom practice, where curriculum, assessment and pedagogy meet around student learning.

The teaching challenge is how schools help students develop the knowledge, judgement, confidence, and capabilities they need in a world where AI is increasingly available.

Central to achieving this is the teacher–student relationship. Compassion, high expectations, knowing students, and building belonging are important human elements in an AI-rich environment. AI may assist aspects of teaching and learning, but it cannot replace the work of building trust, noticing students, and knowing when to support, challenge, or step back.

The place of AI in mathematics classrooms should therefore be considered not only in terms of the quality of its output, but also in terms of the quality of the mathematical thinking it supports.

Staying mathematically in charge

There are some genuine opportunities here. For students, AI can make learning more accessible by scaffolding task entry, providing additional explanations, generating varied examples/practice questions, as well as giving timely feedback. This can increase students’ confidence, motivation, and sense that mathematical ideas are within their reach.

At the same time, this support needs careful handling. The risk of cognitive outsourcing is real, where students using AI for the thinking and decision-making that should remain part of their mathematical learning. A useful classroom principle is to have students think first, then use AI to test, extend, or refine that thinking. This matters most when students are still building the knowledge and confidence needed to do the work themselves.

Similarly, AI can assist teachers with designing more engaging contexts, planning, meeting the differing needs of learners, and generating tasks. However, if core mathematical and pedagogical decisions are increasingly outsourced, teacher agency and professional capability are likely to be diminished.

The issue is not whether outsourcing is inherently desirable or not; some outsourcing is part of effective tool use. The real pedagogical questions are therefore:

• What is being outsourced?

• When is it being outsourced?

• Why is it being outsourced?

• Who is responsible?

AI can quickly produce work that looks polished. However, learning is not the same as output. The messy process of trying, revising, explaining, getting stuck, and finding a way through is often where the learning occurs.

That does not mean AI should necessarily be excluded; it means that its classroom use needs to be deliberate.

Strengthening capabilities for mathematical agency

Across the series, several capabilities have emerged as increasingly important in an AI-active world: posing good questions, noticing assumptions, comparing methods, explaining why a result is reasonable or not, and identifying what can be relied on and what needs further checking. In classroom practice, these capabilities need to be deliberately built into the work students are asked to do.

A useful lens for thinking about classroom use is: What mathematical action is the student taking? If AI use makes that action clearer, more accessible, or more open to discussion, it may strengthen learning. If it removes the need or opportunity for students to take that action, it is more likely to weaken it.

How AI is being used now

In practice, AI use in mathematics classrooms is still developing. It differs across classes and schools, shaped by access, teacher confidence, policy settings, and the guidance available to teachers and students.

At present, much of teachers’ AI use centres on planning and preparation. Students are also using AI across more learning contexts, with classroom use generally more structured than self-directed use outside school. Many purpose-built education platforms are designed to act like Socratic tutors: not simply giving answers but asking probing questions, or offering hints so that students remain active in the thinking.

In primary mathematics classrooms, AI use is generally bounded and teacher mediated. Where it is used, it is mainly in the background through learning platforms that help identify where students are in their learning, provide hints, vary representations, or alert teachers to possible errors or misconceptions.

In secondary mathematics classrooms, student use tends to focus on task and language clarification, and feedback on their work. Students may also work with AI assistants, such as custom GPTs or Gems, created by teachers for particular task- or topic-specific learning activities.

From support to purposeful use

These forms of support play an important role; however, purposeful use asks something more of the task. AI can help bring students into mathematical situations that would otherwise be harder to explore because of their complexity and the time they take to work through.

Purposeful use begins with teachers creating tasks that carry richer mathematical demand. These tasks might involve messy data, larger data sets, different assumptions within the same model, competing approaches, or investigations of how changing a condition affects the result.

These tasks require students to work with greater uncertainty, more variation and less pre-packaged structure. The teacher’s role is to build this kind of work into the learning, while providing suitable structure for students to engage productively. This means deliberately teaching the mathematical investigative skills and the AI-use skills that support them: how to begin with a partial idea, ask a workable question, vary conditions, look for structure, move between representations, and check whether generated results make mathematical sense.

Used this way, AI shifts attention towards the decisions, structures, and relationships that students need to understand.

Teacher knowledge matters

AI does not reduce the need for teacher knowledge.

It offers new possibilities for planning, explanation, tutoring, feedback, and differentiation; but those possibilities do not unfold themselves. Effective implementation depends on teachers deciding what is mathematically worthwhile, pedagogically sound, and suitable for students at different stages of schooling.

In practice, teachers need to know the mathematics itself, understand how it is structured across the curriculum, and be able to teach it in ways that address representations, misconceptions, explanation, and progression.

AI can generate material, but teachers still need to judge its mathematical quality, purpose, timing, and fitness for learning.

System support is vital

If schools are to implement AI thoughtfully in mathematics, they will need more than access to tools and broad encouragement. They will need sample tasks, examples of practice, subject-specific guidance, and professional learning that connects curriculum, pedagogy, assessment, and ethical use.

Implementation is rarely just about introducing something new. It depends on professional confidence, capability, shared expectations, and the conditions that enable teachers to use AI well. It also needs to reduce variation within and between schools.

This points to a system responsibility.

Australian systems and sectors have started exploring the use of generative AI in different ways. ACARA has developed a ‘Curriculum Connection: artificial intelligence’ resource to map AI to learning areas. Some state and territory departments and Catholic school sectors have also developed bounded AI platforms for safer and more structured school use, including in mathematics. Examples include EdChat (SA), EduChat (NSW), and cechat (Catholic Education).

Final thought

Classroom practice is where curriculum, assessment, and pedagogy come together around student learning. The educational value of AI in school mathematics lies in whether it helps students think more deeply, participate more fully, and show more clearly what they understand. AI use should be judged by the mathematical actions it enables or removes.

So, the concluding question for this series is this:

How will schools and systems use AI in ways that deepen mathematical learning, protect student agency, and ensure that AI-supported output is not regarded as a proxy for mathematical understanding?

AI statement:

This blog post was written by a human. ChatGPT was used for final editorial proofreading before being reviewed by a human editor.

References

This series has drawn on a range of research papers, policy reports, professional statements, and practitioner-facing materials. Together, these sources informed various elements of the curriculum: mathematical, assessment, pedagogical, and governance questions raised by AI in school mathematics. The list below is selective rather than exhaustive. It is intended as a short background reading set for readers who would like to pursue the issues further.

Alfarwan, A (2025). Generative AI use in K–12 education: A systematic review. Frontiers in Education.

Australian Government Department of Education. (2025). Australian framework for generative artificial intelligence (AI) in schools. Australian Government.

Castlereagh Summit (2026) The Castlereagh Statement: A cross-sector call to action on Australian education and training in the age of AI. Sydney, Australia.

Eyal, L. (2025). Developing and validating an AI-TPACK assessment framework: Enhancing teacher educators’ professional practice through authentic artifacts. Education Sciences, 15(11), 1452.

Fesler, L., Martinez, J., Agnew, C., Loeb, S. (2026). The Evidence Base on AI in K-12: A 2026 Review, AI Hub for Education of the SCALE Initiative, Stanford University.

Fitzpatrick, D. (Host). (2025–2026). AI for Educators Daily with Dan Fitzpatrick [Podcast]. Apple Podcasts.

Fitzpatrick, D. (2026, January 30). 8 AI research papers you should read in 2026. The AI Educator.

Hussein, H. B. (2026). Artificial intelligence in mathematics education: A systematic review. International Journal of Research in Educational Sciences, 9(2), 447–478.

Lodge, J. M., & Loble, L. (2026). Artificial intelligence, cognitive offloading and implications for education. Network for Quality Digital Education / University of Technology Sydney.

Maksimchuk, M. T., et al. (2025). Generative AI in the K–12 formative assessment process: Enhancing Feedback in the Classroom. In Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con) volume 1, 107–110.

Mishra, P., & Koehler, M. J. (2006). Technological pedagogical content knowledge: A framework for teacher knowledge. Teachers College Record, 108(6), 1017–1054.

National Center on Education and the Economy. (2026, March 26). Teaching, learning, and leading in an AI-augmented world. NCEE.

National Council of Teachers of Mathematics. (2024). Artificial intelligence and mathematics teaching [Position statement]. NCTM.

National Council of Teachers of Mathematics. (2024). High school mathematics reimagined, revitalized, and relevant: Executive summary. NCTM.

OECD (2025). Evolving AI capabilities and the school curriculum: Emerging implications and a case study on writing.

OECD. (2025). What should teachers teach and students learn in a future of powerful AI?. OECD.

OECD. (2026). OECD digital education outlook 2026: Exploring effective uses of generative AI in education. OECD Publishing, Paris.

Puentedura, R. R. (2014, June 29). Learning, technology, and the SAMR model: Goals, processes, and practice. Hippasus.

Son, T.(2024). Intelligent tutoring systems in mathematics education: A systematic literature review using the Substitution, Augmentation, Modification, Redefinition model. Computers, 13(10), Article 270.

UNESCO. (2023, updated 2026). Guidance for generative AI in education and research. UNESCO.

UNESCO. (2025). AI and education: Protecting the rights of learners. UNESCO.

University of Melbourne, Faculty of Education. (2025, May 19). Navigating AI in mathematics education: Insights for all educators [Podcast episode]. Talking Teaching.