Everyone's Talking About AI in Secondary Schools. What About Primary?

When you picture the future of AI in primary education, do you see a class of five-year-olds staring at an adaptive tablet app, or do you see a teacher freed from 28 hours of weekly administration? How we answer that question right now will define the next decade of primary pedagogy.

Generative AI has burst onto the education scene in recent years, seemingly with few limits or clear guidance on how we should be using it. The media debates that are raging around its use in education predominantly reflect a secondary and further education perspective of plagiarism, essay writing and revision. What about our primary colleagues? How can the use of generative AI enhance this sector and what aims should we be striving for to ensure that education in primary schools today doesn't lose sight of the uniqueness of how children develop, learn and thrive?

The Admin Burden: Why Teachers Are Burning Out

Teachers are overwhelmed and have been for many years. They are expected to introduce more and more into their classrooms, many of which are at odds with one another. How do you ensure that all pupils move more, talk more and collaborate more whilst also sitting still, working alone and passing exams? Not to mention all the additional admin you need to produce to show you have achieved this, or not.

This is not conjecture; it is a fact backed up by report after report. As the DfE 'Working Lives of Teachers and Leaders' survey shows, a full-time teacher works over 51 hours per week during term time, with those in leadership reporting nearly 57 hours per week. Yet, with less than half that time spent directly educating children, no wonder the admin burden is becoming untenable for teachers.

In primary education, it is an aim for every teacher to be able to plan interesting lessons with engaging resources, personalised to all the diverse needs of the class and then have the energy left to deliver the lesson. It is totally possible, if you only worked in a classroom half the week.

What if we can harness generative AI to do the heavy lifting of this admin in a fraction of the time? Teachers get their own personal assistant to support planning, resourcing and reporting so they can focus the majority of their time on building connections with pupils. The Education Endowment Foundation (EEF) has demonstrated that when generative AI is used for lesson preparation, it can reduce lesson planning time by 31%.

The pupils perspective: The Human Experience

Let's now consider how children fit into this AI landscape. The late Sir Ken Robinson once posed the question:

"What is it about us, and our children, that a machine cannot replicate?"

This is even more crucial to consider today in light of the rapid advances in computing we have seen through generative AI. Generative AI 'pretends' to be creative and can make lovely images, videos and write stories using complex grammatical structures that hit every curriculum checkbox. If we took AI to its extreme, we could ask a five-year-old to say 'draw me a picture of a cat' into an AI tool and accept the response as ticking the box for that child's 'art skills'. Why would we bother with learning how to draw ourselves if AI can do it all?

The human experience. That's why.

Humans are unique and we need to embrace the idea that our brains and bodies are stimulated and enhanced through exploring, experiencing awe and wonder or imagining possibilities that could take us anywhere. This is why the potential of generative AI in our education system needs to be considered with this always in mind.

AI as a Launchpad, Not the Destination

For primary pupils, generative AI must never be the end product; it must be the launchpad. Consider the traditional, often paralysing classroom scenario: a child is presented with a stark, blank piece of paper and told to "write a story". For many, the cognitive load of generating ideas, recalling vocabulary and physically spelling words creates an immediate wall.

Now, imagine a classroom where AI is leveraged to enable the pupils' creativity to flow. Rather than staring at a blank page, the child collaborates with AI to brainstorm, prompting it to generate a rich, vivid list of adjectives based on their chosen topic. The child then inputs those exact adjectives into an AI image generator. Their abstract vocabulary is converted into a vibrant visual image. They aren't looking at a generic illustration; they are looking at a world they built with their own words. Finally, in the excitement of their image, the child tells their story aloud. A natural language AI transcribes their voice in real time, capturing every enthusiastic tangent, every 'um', 'ah' and repetition.

The true creative impact of this happens when the child looks at the screen. They are not looking at a piece of work a chatbot wrote for them; they are looking at their own authentic voice, captured and made visible. The gruelling physical barrier of transcription was handled by the technology, leaving the child with the cognitive energy to continue engaging in their creation and a tangible starting point for editing. They can actively collaborate with the teacher to spot the 'umms', recognise where a sentence repeats, swap out words for better alternatives and experiment with punctuation.

By shifting AI to the very beginning of the creative cycle, we replace the intimidation of the blank page with an empowered editing process. We aren't bypassing the hard work of learning to write; we are giving children the scaffold they need to become confident, critical editors of their own extraordinary ideas.

Personalised Learning: Supporting SEND pupils

Now let's think about a SEND child who is obsessed with fire engines. As educators, we may not know everything there is to know about Fireman Sam, but generative AI certainly can. It can tell you how to teach doubling or adverbs, or even the water cycle using this niche topic. It can be the spark that ignites that love of learning for that particular child. We are on the cusp of having the chance for truly personalised learning, which research has repeatedly told us gives the best outcomes for our young people. Crucially, the opportunity to have it without an untenable admin burden associated with it.

New AI-powered apps are arriving in our schools every day, of varying qualities and uses. Some tools are amazing, allowing children to navigate their own learning at a structured pace, supporting diverse SEND needs with text levelling software and eye movement sensors for pupils with physical impairments. However, as the CEO of Khan Academy recently said about their AI learning chatbot, it 'doesn't necessarily make students motivated to learn or fill in gaps in knowledge needed to ask questions.' (as reported by Chalkbeat).

The Future of Primary Education 

This is the message we need to reflect on when adopting any new software, but especially with artificial intelligence. Does it actually enhance learning or is it more a hindrance than a help? If it's the latter, the message for school leaders must always be to rethink and never lose sight of professional discretion.Generative AI isn't going away and education does need to understand and embrace its potential. For school leaders and teachers, it offers the release from hours of paperwork and the streamlining of systems. Whilst it does have the potential to enhance professional skills, we should not be expecting our over-burdened teachers to just teach themselves. There must be training available, in contracted hours, to ensure this new technology can truly be revolutionary and not just another stress to be added.

For pupils in schools today, it can open the door to new and exciting ways of learning, in collaboration with educators, never replacing them. Our young children need to grow up understanding the benefits of AI but also, most crucially, the limitations of it. Through harnessing this approach, schools can continue to be a place of imagination, connection and creativity, but this time with a workforce that isn't facing admin burnout.

References

Department for Education. (2025). Working Lives of Teachers and Leaders: Wave 4. Available at: https://www.gov.uk/government/publications/working-lives-of-teachers-and-leaders-wave-4

Education Endowment Foundation. (2024). Teachers using ChatGPT – alongside a guide to support them to use it effectively – can cut lesson planning time by over 30 per cent. Available at: https://educationendowmentfoundation.org.uk/news/teachers-using-chatgpt-alongside-a-guide-to-support-them-to-use-it-effectively-can-cut-lesson-planning-time-by-over-30-per-cent

Robinson, K. (2019, September). Creativity can change the world [Speech]. Energy Disruptors UNITE (EDU2019), Calgary, AB, Canada.

Barnum, M. (2026). Sal Khan once said AI would revolutionize schools. Now he sees its limits. Chalkbeat, 9 April 2026. Available at: https://www.chalkbeat.org/