Can AI Really Personalise Learning for Every Child?

AI promises to personalise learning for every child. Some of that promise is real. Some of it is overstated. Here's the honest version of what AI can and can't do in a primary school classroom.

What "personalised learning" actually means — and what AI has to do with it

Personalised learning is one of those phrases that education has been promising for decades and rarely fully delivering on. The idea — that every child learns differently, so every child should have learning that adapts to them — is obviously right. The challenge has always been that one teacher with 30 pupils can only personalise so much.

AI changes some of that. But it's worth being precise about which parts it changes and which parts it doesn't.

What AI can genuinely do for individual learners

Some AI tools in education are designed specifically to adapt to a learner's level, pace and responses. Oak National Academy's AI tool Aila, for example, generates lesson content structured around specific year groups and learning objectives. Tools like Curipod create interactive activities that respond to pupil input in real time.

Beyond those direct learning tools, AI helps teachers personalise in less obvious but equally important ways:

• Differentiated resources — a teacher can ask an AI tool to produce the same activity at three different reading levels in minutes, rather than hours. Children with SEND or those working above expected levels get materials that actually suit them.
• Faster feedback — when marking and planning take less time, teachers have more time to give individual children the feedback that actually moves their learning forward.
• Comment banks for reports — AI helps teachers produce the structural framework for reports, freeing up time to write the genuinely personalised observations that parents most want to read.

What AI cannot personalise — and why that matters

Here's the important counterpoint. The most powerful form of personalised learning isn't algorithmic — it's relational. It's a teacher who knows that one child needs encouragement before attempting anything new, and another needs to be challenged before they'll try. It's noticing that a pupil who was engaged last week seems flat today and following up. It's adjusting the energy of a lesson because you can read the room.

No AI tool can do any of those things. They require a human being who knows the child — and that remains the most important ingredient in good teaching, full stop.

What this means for primary-age children specifically

For younger children especially, the research on personalised learning points strongly towards the relationship with a trusted adult as the most significant factor in learning outcomes. Children learn best when they feel safe, seen and supported by their teacher. No software changes that equation.

Where AI adds genuine value at primary level is in reducing the time teachers spend on tasks that don't require their professional judgement — which gives them more time and energy for the relational, responsive teaching that actually personalises learning in the deepest sense.

Questions to ask your child's school

If personalised learning is important to you as a parent, the most useful thing to understand is how your school uses assessment data to inform teaching — and whether AI tools are supporting that or getting in the way of it. A school that has adopted AI thoughtfully will be able to give you a clear answer. If you want to know more about how your school approaches AI, our guide for parents is a good starting point.