How UK Teachers Can Use AI for Creating Presentations

A Year 4 teacher planning a fractions unit, a Key Stage 3 geography teacher covering plate tectonics, and a Reception practitioner introducing a new phonics phase all share one recurring, unglamorous task: turning subject knowledge into a slide deck someone else can follow. Building a presentation is rarely the hard intellectual work of teaching — the hard work is deciding what to teach and how to sequence it. But the mechanical build-out still eats real hours every week across English classrooms.

That mismatch is exactly where AI tools are proving useful for teachers in England. Not by deciding what to teach — that decision has to stay with the person who knows the class, the programme of study, and where each child currently sits against it — but by handling the layout, the drafting, and the differentiated variants that used to mean an evening at the laptop.

This guide walks through what the National Curriculum actually expects a presentation to do at different key stages, where AI genuinely earns its place in that workflow, and where it falls short. It also covers the UK-specific data protection questions any school should ask before rolling a tool out.

What the National Curriculum Actually Expects From a Lesson Presentation

England's National Curriculum does not specify slide formats — it specifies programmes of study and, at the end of key stages, attainment targets that pupils should be working towards. A presentation is only a vehicle for delivering that content; it is judged (by inspectors, by mentors, by the teacher's own evaluation) on whether it supports pupils in meeting those targets, not on how polished it looks.

Key stage expectations differ more than people assume

• EYFS (Reception, age 4–5): the curriculum is built around the seven areas of learning, delivered largely through play and adult-led activities rather than didactic slides. A "presentation" here is closer to a visual prompt — a picture book cover, a small set of images to anchor a carpet-time discussion — not a multi-slide deck.
• KS1 (Years 1–2, age 5–7): phonics screening in Year 1 and the broader KS1 programmes of study mean visuals need to be simple, high-contrast, and text-light, supporting oral explanation rather than replacing it.
• KS2 (Years 3–6, age 7–11): pupils can handle denser content and are building towards the Year 6 SATs, so presentations can carry more structured information, worked examples, and diagrams tied to specific attainment targets in maths, English, and science.
• KS3 (Years 7–9, age 11–14): subject specialism sharpens. History, geography, and science presentations need to reflect the discipline's own conventions — sourcing and interpretation in history, fieldwork and data in geography, the scientific method in science — because pupils are being prepared for GCSE study.

The presentation has to reflect the sequence of learning, not just the topic

A common mistake is treating a presentation as a standalone artefact when it is really one slice of a sequenced unit. The National Curriculum's programmes of study are written as progressions — what pupils build on from the previous key stage — so a strong presentation makes explicit links backward (what was covered before) and forward (what this sets up). If you're covering Year 5 history, for instance, a presentation that ignores what pupils met in Year 3 and 4 will feel disconnected even if the slide design is excellent. Our related guide on AI tools for Year 5 history in the UK goes deeper on sequencing history content specifically.

Assessment objectives should be visible, not implicit

For KS2 and KS3 in particular, a presentation that supports formative assessment tends to work better than one that is purely informational. That means:

• A visible learning objective slide, phrased against the relevant programme of study strand
• At least one slide built around a check-for-understanding question, not just content delivery
• A closing slide that connects back to the objective, rather than simply ending on the last content point

Where AI Genuinely Helps — and Where It Doesn't

It's worth being blunt about the boundary here, because overclaiming what AI can do for lesson presentations is how classrooms end up with generic, curriculum-mismatched slides.

What AI tools are actually good at

• Structuring a rough outline into slides. Given a topic, key stage, and learning objective, an AI tool can propose a sensible slide sequence (starter, explanation, worked example, practice, plenary) far faster than building one from a blank template.
• Drafting differentiated text variants. Producing a simplified version of an explanation for pupils working below age-related expectations, and an extension version for greater-depth pupils, from the same base content.
• Generating supporting materials in the same session. Question sets, vocabulary glossaries, or a matching worksheet that reinforces the same slide content, so the presentation and the accompanying task are aligned rather than separately sourced.
• Speeding up visual consistency. Suggesting a consistent slide layout and heading style across a long unit, which is tedious to maintain by hand across 30+ slides.

What AI tools are not good at — and shouldn't be trusted with

• Curriculum judgement. An AI tool has no knowledge of your specific class, their prior attainment, or where the school is in its own curriculum sequence. It can draft content aligned to a general description of a key stage, but a teacher still has to check it against the actual programme of study and the school's own scheme of work.
• Subject accuracy at the edges. Generated content can be confidently wrong, especially with dates, named events, or scientific specifics. Every factual claim on a generated slide needs a teacher's own subject knowledge applied before it reaches pupils.
• Sensitivity and safeguarding judgement. Topics that touch on sensitive family circumstances, bereavement, or safeguarding-adjacent themes need a human's judgement about tone and framing — not an AI-drafted default.
• SEND and EAL differentiation nuance. AI can produce a simplified variant on request, but it does not know an individual pupil's specific access needs, EHCP targets, or the exact scaffolds a TA already uses with that child.

The honest framing: AI tools can compress the build-out phase of presentation-making from hours to minutes. They cannot compress the professional judgement phase — deciding what's worth teaching, in what order, and how to check it landed. That still belongs entirely to the teacher.

Practical AI Workflows for Building Lesson Presentations

A workflow that keeps the teacher in control

A reliable pattern many teachers report finding useful is a three-step loop rather than a single "generate a presentation" prompt:

1. Draft the skeleton yourself first — the learning objective, key stage, and the two or three big ideas the lesson has to land. This takes two minutes and keeps the AI output anchored to your actual plan rather than a generic version of the topic.
2. Ask the AI tool to expand the skeleton into slide content, specifying key stage, subject, and any differentiation you need (e.g. "one simplified explanation slide, one extension slide").
3. Review and adjust every slide against the programme of study strand it's meant to serve, correcting anything factually off and rewording anything that doesn't match how your class talks about the topic.

Platforms designed for teachers can help at step two specifically. EduGenius, for example, is designed to generate lesson materials — including slide content, worksheets, flashcards, and mind maps — from a class profile that reflects the grade and ability level you specify, then export the result to PPTX, PDF, or DOCX for editing in your usual tools.

Prompt ideas by key stage

Specific, well-scoped prompts tend to produce far more usable output than vague ones. Some starting points:

• KS1 (Year 2, science — everyday materials): "Draft five simple slides introducing the properties of everyday materials for Year 2, with one image-led slide and short, high-contrast text suitable for early readers."
• KS2 (Year 5, maths — fractions): "Draft a slide sequence explaining equivalent fractions for Year 5, including one worked example slide and one slide with three practice questions of increasing difficulty."
• KS3 (Year 8, geography — plate tectonics): "Draft slides explaining plate boundary types for Year 8 geography, including a labelled diagram description and one slide connecting the topic to a real, named example of tectonic activity."
• KS2 (Year 6, English — persuasive writing): "Draft a slide deck modelling the structure of a persuasive letter for Year 6, with one annotated exemplar slide showing rhetorical devices."

Building differentiated variants without rebuilding from scratch

Rather than writing three separate presentations for a mixed-ability class, many teachers draft one core deck and then ask an AI tool for two variants of the explanation slides only — a scaffolded version and an extension version — which are then swapped in for the relevant groups. This keeps the shared structure (objective, plenary, practice questions) consistent across the class while adjusting only the slides where ability actually matters.

Pairing presentations with other AI-generated materials

A presentation rarely stands alone in a real lesson — it's usually paired with a worksheet, a set of questions, or flashcards for retrieval practice. Generating these from the same source content, in the same sitting, keeps vocabulary and framing consistent across the lesson's materials. If your unit also needs quick recall practice, the approach used in our guide on how UAE teachers can use AI for making flashcards translates directly to a UK classroom, and pairs well with slide-based instruction.

A comparison of AI-tool categories for presentation building

Choosing Tools Responsibly: Data Privacy in an English School

Data protection is not a side issue when adopting AI tools in a UK school — it's a compliance requirement, and it's one area where cutting corners creates real institutional risk.

The UK GDPR baseline

Any tool that processes pupil data — names, class groupings, attainment data, or anything that could identify a child — falls under UK GDPR and the Data Protection Act 2018. The Information Commissioner's Office (ICO) publishes UK GDPR guidance for organisations, including material relevant to schools, and it's worth checking a tool's data-handling terms against that guidance before adopting it school-wide, not just at classroom level.

Practical steps before adopting a tool

• Check whether the tool needs pupil-identifiable data at all. For presentation-building specifically, most workflows only need the topic, key stage, and subject — not pupil names or attainment records — so choosing a tool that doesn't require personal data sidesteps a lot of the compliance burden.
• Confirm where data is processed and stored, and whether the vendor's terms meet the school's own data protection policy.
• Follow your school's own AI usage policy, if one exists, or raise the gap with your data protection lead if it doesn't yet cover generative AI tools.
• Review the Department for Education's published guidance on generative AI in education, which sets out expectations for schools adopting these tools safely.

Keep pupil work out of prompts by default

A simple habit avoids most privacy questions entirely: when drafting a presentation, describe the topic and key stage, not the pupils. "Draft a Year 4 fractions explanation slide" carries no personal data risk. "Draft a slide explaining fractions for a pupil who struggled with the topic last week" starts to edge toward pupil-specific detail — and that kind of detail almost never needs to go into an AI prompt to get a useful result.