How UK Teachers Can Use AI for Planning Lessons

Planning in England has never been a matter of filling in a template. Every lesson has to trace back to the National Curriculum's programmes of study, sit inside a department's or key stage's sequencing, and account for a room full of children working at different points along that sequence. Add statutory assessment points — the Year 1 phonics screening check, Key Stage 2 SATs — and it is easy to see why planning time is one of the first places teachers look for relief.

AI tools will not write your curriculum for you, and they should not. But used with a clear head about what they are good at, they can take a real bite out of the repetitive parts of planning — drafting resource variants, generating retrieval questions, building differentiated worksheets — leaving more of a teacher's attention for the judgement calls only a teacher can make.

This guide walks through what planning actually involves at each key stage in England, where AI genuinely helps, where it does not, and how to build a workflow that keeps a teacher firmly in charge of the final product.

What Lesson Planning Actually Demands in England

Planning a lesson well is not one task — it is several overlapping ones, done under real time pressure and against a fixed curriculum framework.

The National Curriculum sets the destination, not the route

Every subject in the National Curriculum is written as a programme of study: a set of statutory attainment targets for the key stage, not a lesson-by-lesson script. That is deliberate. It hands schools and teachers the job of sequencing content, choosing resources, and deciding pace.

Two schools teaching the same Year 4 objective on fractions can arrive at it through entirely different routes — one via a mastery scheme with daily fluency practice, another through a topic-based approach that revisits fractions across several subjects. Both are compliant with the same statutory wording. That flexibility is valuable, but it also means every teacher effectively re-does a chunk of the design work each time, rather than following a single national script.

In practice, a single lesson has to:

• Connect to a specific point in the programme of study for that subject and key stage
• Build on what was taught in the previous lesson (or half-term, or year)
• Offer a route in for pupils below age-related expectations and stretch for those above it
• Produce something that shows evidence of learning, whether that's exercise-book work, a discussion, or a quiz

Workload is the honest starting point for this conversation

The Department for Education's own Teacher Workload Survey has, across several rounds, found that full-time teachers in England report working close to 50 hours a week in term time — and that planning and resource preparation make up a significant share of the hours spent outside the classroom. That is the backdrop against which any "AI for planning" conversation has to be honest: the goal is not novelty, it is time.

A realistic aim for AI in this context is narrower than "planning done for you." It's freeing up the hour a teacher would otherwise spend retyping a worksheet in three ability tiers, so that hour can go toward things AI cannot do — moderating books, talking to a TA about a specific pupil, or rethinking a sequence that clearly isn't landing.

Planning time in most schools breaks down into a handful of recurring, largely mechanical tasks:

• Reformatting the same core content into two or three ability tiers
• Sourcing or creating images, diagrams, and reading extracts to match a topic
• Writing retrieval-practice questions that recap previous learning
• Producing an answer key or mark scheme alongside a resource
• Adapting a resource built for one class to fit a parallel class with a different starting point

None of these tasks require a fresh curriculum decision each time — they require a teacher's earlier decision to be executed multiple times, in slightly different forms. That repetition is exactly where AI tools have the most to offer.

What Good Planning Looks Like Across Key Stages

"AI for lesson planning" means something different in a Reception classroom than it does in a Year 9 history lesson. The framework, the pace, and the risks all shift.

EYFS and KS1: play, phonics, and short attention spans

In Reception (EYFS) and Years 1–2 (KS1), planning is shaped by the Early Years Foundation Stage framework and then, from Year 1, the phased demands of systematic synthetic phonics ahead of the Year 1 phonics screening check. Content has to stay concrete, oral, and playful — a five-year-old does not benefit from a worksheet dense with instructions.

Here, AI is most useful for:

• Generating short, decodable sentence sets pitched at a specific phonics phase
• Producing simple picture-matching or sorting activities for continuous provision
• Drafting parent-facing phonics practice sheets that mirror the school's chosen scheme's tone

KS2: programmes of study meet SATs pressure

Years 3–6 is where subject content visibly deepens — fractions, grammar terminology, longer historical narratives — and where Key Stage 2 SATs in Year 6 add a layer of test-format familiarity to the planning job. A Year 5 teacher planning a unit on fractions, for instance, has to hold both the programme of study wording and the reality that pupils arrive at wildly different starting points.

AI can help by generating:

• Multiple worked-example sets at different difficulty tiers for the same objective
• SATs-style question banks in the correct format (reasoning papers, arithmetic-style layout)
• Retrieval-practice starters that recap the previous half-term's content

The stakes are slightly different in Year 6, where every lesson from the autumn term onward is implicitly building toward a set of statutory tests. A teacher planning a Year 6 grammar lesson has to balance genuine understanding of a concept against the very specific way SATs papers phrase questions about it — two things that don't always line up neatly, and that a teacher's professional judgement has to reconcile.

KS3: subject depth and the runway to GCSE

Years 7–9 sit inside subject departments where planning is increasingly about depth and disciplinary skills — analysing a source in history, structuring an argument in English, applying a method in maths — as a foundation for GCSE options. Planning here often means adapting a scheme of work written by someone else (a head of department, a previous cohort's teacher) to a specific class.

AI is genuinely strong at:

• Turning a scheme-of-work objective into a starter, main task, and plenary in minutes
• Producing differentiated reading extracts or source material at several reading ages
• Drafting exam-style questions with mark schemes for formative use

Where AI Genuinely Helps With Planning — and Where It Doesn't

A useful mental model: AI is good at producing variants of something a teacher has already decided the shape of. It is weak at deciding what should be taught, and whether a pupil actually understood it.

Strong use cases

• Differentiation at speed. Turning one core resource into three or four tiered versions is exactly the kind of repetitive, rules-based task AI handles well.
• First-draft resource generation. Worksheets, flashcards, mind maps, and starter quizzes can be drafted in a fraction of the time it takes to build from scratch.
• Retrieval practice. Short recap quizzes tied to a specific prior topic are quick to generate and easy for a teacher to verify against the scheme of work.
• Explaining a concept three different ways. Useful for finding an alternative explanation for a pupil who didn't grasp the first one.

Where it falls short

• Curriculum sequencing judgement. Deciding what a class needs next, based on how yesterday's lesson actually went, is not something AI observed.
• SEND and EAL nuance. A generated "differentiated" resource still needs a teacher's eye for a specific pupil's needs — AI has no visibility into an EHCP or an individual pupil's history.
• Subject accuracy at the edges. Any generated content — a historical date, a scientific explanation, a grammar rule — needs a subject specialist's check before it reaches pupils.
• Marking and moderation. AI can draft a mark scheme; it cannot make the professional judgement call that moderation meetings exist for.

The rule of thumb worth keeping on a sticky note above the desk: AI drafts, the teacher decides.

It also helps to be honest about a subtler risk: over-reliance can quietly erode a teacher's own subject and pedagogical fluency over time, especially for early-career teachers still building that expertise. Treating AI output as a first draft to be interrogated — rather than a finished answer to be accepted — keeps that skill-building intact rather than short-circuiting it.

Practical AI Workflows and Prompt Ideas by Key Stage

Vague prompts produce vague, unusable output. The more a prompt mirrors how a teacher would brief a very literal-minded trainee, the better the result — key stage, subject, objective, and constraints, spelled out every time.

Prompt structure that works

A dependable prompt for planning tends to include, in order:

1. Key stage and year group ("Year 4", "KS3 Year 8")
2. Subject and specific objective, ideally quoting the programme of study wording
3. Format wanted (worksheet, quiz, starter, mind map)
4. Differentiation need (e.g., "three tiers: working towards, expected, greater depth")
5. Length or time constraint ("a 10-minute starter", "one side of A4")

Worked examples

• KS1 phonics: "Generate a set of 10 CVC decodable sentences using only Phase 3 graphemes, for a Year 1 guided reading group, with three sentences per page and large print."
• KS2 maths: "Create a differentiated worksheet on adding fractions with different denominators for Year 5, in three tiers, following the National Curriculum wording for that objective, with an answer key."
• KS3 history: "Write three source-analysis questions on the causes of the First World War for Year 9, with a mark scheme, pitched at a mixed-ability class."

This is precisely the gap that platforms like EduGenius are designed to help with — the tool can generate worksheets, quizzes, flashcards, and mind maps against a stated year group and objective, with an answer key attached, so the teacher's editing time goes toward checking accuracy and fit rather than formatting from a blank page. Class profiles that remember a group's typical ability spread can also make repeat requests faster to brief.

A key-stage-by-use-case view

Choosing Tools Responsibly: Data and Quality Control

None of this is worth doing if it creates a data-protection problem or quietly lowers the accuracy bar in the classroom.

UK GDPR and DfE guidance on generative AI

Any tool that touches pupil data in England has to be considered against UK GDPR and the school's own data protection policy — the Information Commissioner's Office publishes guidance on what UK GDPR requires of organisations processing personal data, which applies to schools too. The Department for Education has also published guidance for schools on the safe, effective use of generative AI, which is worth reading alongside a school's own AI policy before rolling out any tool.

Practical habits that keep planning-with-AI on the right side of that guidance:

• Never paste identifiable pupil information — names, SEND status, EHCP detail — into a general-purpose AI tool
• Check what a tool's terms say about training on submitted data before using it with anything sensitive
• Favour tools built for education that are explicit about data handling, over general consumer chatbots
• Keep AI use for planning and resource creation, not for storing or processing pupil records

The quality-check habit that has to stay non-negotiable

Every AI-generated resource needs the same check a teacher would give a colleague's handout: read it fully, check it against the programme of study wording, and confirm the difficulty tiers actually match the class. This is not optional extra work bolted onto AI use — it's the professional judgement step that makes AI-assisted planning safe to use at all.