How UK Teachers Can Use AI for Writing IEP Goals
UK teachers can use AI to draft SMART, measurable individual education plan (IEP) targets faster by feeding it a student's current attainment, the specific area of need, and the SEND Code of Practice outcome category, then editing the draft against what the teacher and the child's family actually know about that pupil. AI drafts the language quickly; the teacher still owns every judgment call about what the target should be and whether it's realistic.
Quick Answer: Use AI to turn a rough description of a pupil's need into a well-worded, measurable target aligned to the SEND Code of Practice's four broad areas of need, then review it against the pupil's actual data and your professional judgment before it goes in the plan. AI speeds up the writing; it never decides what a child needs.
England doesn't run a single national "IEP" system the way the term is used in the US — since the 2014 Children and Families Act, statutory support runs through SEN Support and, for more complex needs, an Education, Health and Care Plan (EHCP). In practice, though, many schools still keep an internal working document, often still called an individual education plan (IEP) or a "SEN support plan," to track a pupil's specific targets term by term — and that's the document this guide is about.
This article covers:
- Why writing measurable IEP/SEN Support targets is such a time-heavy task
- What AI is genuinely good at in this specific writing task, and where it isn't
- A step-by-step workflow for drafting a target with AI
- How to keep AI-assisted targets compliant with the SEND Code of Practice
- Common mistakes that undermine an otherwise well-drafted target
Why Writing IEP and SEN Support Targets Takes So Long
Writing a genuinely SMART target — specific, measurable, achievable, relevant, and time-bound — for each pupil on a SEN register is a skilled, time-consuming task, and most SENCOs (Special Educational Needs Coordinators) are managing this across dozens of pupils at once, not just one or two. The Education Endowment Foundation (EEF) has consistently found in its guidance on special educational needs that specific, measurable targets are strongly linked to better outcomes than vague ones — which raises the writing bar exactly where teacher time is already stretched thinnest.
The SENCO Time Problem
A single well-written target might take fifteen to twenty minutes to draft properly: describing current attainment accurately, choosing a genuinely measurable success criterion, and setting a realistic timeframe. Multiply that across a SEN register that can run into dozens of pupils per SENCO, each needing termly review, and the writing burden becomes one of the least visible but most demanding parts of the role — work that happens largely outside contact time, on top of an already full timetable.
Where AI Fits Without Replacing Judgment
AI can turn a rough, informal description — "Year 3, struggles with letter reversals in writing, needs support with phonics blending" — into properly worded, measurable target language in seconds. What it can't do is know whether that target is the right one for this specific child, which depends on classroom observation, family input, and often an educational psychologist's assessment that no AI tool has access to.
What AI Can (and Can't) Do When Writing IEP Goals
AI is strongest at producing well-structured, measurable target language from a teacher's informal notes, and weakest at deciding what the actual target content should be for a specific child.
| Task | AI's appropriate role | What still needs the teacher/SENCO |
|---|---|---|
| Turning informal notes into SMART-worded target language | Strong | Confirming the target reflects this specific pupil accurately |
| Suggesting measurable success criteria | Strong as a starting point | Setting a criterion realistic for this pupil's actual progress rate |
| Aligning target wording to SEND Code of Practice language | Strong | Ensuring the underlying need category is correctly identified |
| Deciding what a pupil actually needs support with | Not appropriate — no access to assessment data or classroom context | Teacher, SENCO, and often an educational psychologist |
| Writing family-facing summary language for a review meeting | Useful first draft | Tone and accuracy check before it reaches parents |
| Judging whether a target has been met at review | Not appropriate as sole judge | Teacher's observation and evidence review |
A Step-by-Step Workflow for Drafting an IEP/SEN Support Target with AI
This sequence works whether you're using a general AI assistant or a purpose-built content tool, and it keeps the teacher's judgment at the two points that matter most: the input and the final check.
- Start with what you actually know about the pupil. Note current attainment specifically (not "struggles with reading" but "reads CVC words accurately but reverses b/d in roughly half of attempts"), the broad area of need, and any relevant context from family or previous reviews.
- Ask AI to draft the target in SMART, Code of Practice-aligned language. Specify the area of need category (communication and interaction; cognition and learning; social, emotional and mental health; or sensory and/or physical) so the wording reflects the right framework.
- Request a measurable success criterion, not just a description. Ask specifically for "how progress will be evidenced" — a number, a percentage, or an observable behaviour, not a vague "will improve."
- Check the draft against what you and the family actually know. This is the step that can't be skipped: does the target reflect a realistic timeframe and a genuine priority for this child, not just plausible-sounding language?
- Adjust the wording, not just accept it wholesale. Edit anything that doesn't match your professional judgement about pace, priority, or phrasing your school typically uses.
- Save the final version in the pupil's actual plan, and keep your informal notes as a record of what informed it, useful for the next termly review.
Keeping AI-Assisted Targets Aligned with the SEND Code of Practice
The 2015 SEND Code of Practice sets out four broad areas of need and expects targets to follow an "assess, plan, do, review" cycle — a structure AI can help you word consistently but cannot itself apply, since it doesn't have access to the assessment data or the review evidence that cycle depends on.
The Four Areas of Need
Specifying which of the four broad areas a target falls under produces meaningfully better AI output, because the language and typical target structure differ across them:
- Communication and interaction (including speech, language, and autism spectrum needs)
- Cognition and learning (including specific learning difficulties like dyslexia)
- Social, emotional and mental health
- Sensory and/or physical needs
Where AI Output Needs the Closest Review
Targets touching social, emotional, and mental health needs deserve the most careful human review of the four categories, since AI-generated language here can sound plausible while missing context only a teacher who knows the child would catch — a specific trigger, a family circumstance, or a strategy that's already been tried and didn't work.
Tools UK Teachers Can Use for This Task
Both general AI assistants and purpose-built content platforms can help with target drafting, though they suit slightly different parts of the workflow.
| Tool | Best for | Typical cost | Caution |
|---|---|---|---|
| ChatGPT / Gemini / Claude | Flexible drafting and rewording of informal notes into SMART language | Free tier; paid tiers roughly £16-20/month | Requires you to specify the Code of Practice framework explicitly |
| EduGenius | Generating differentiated worksheets and materials aligned to a pupil's class profile once a target is set | 25 free welcome credits; Starter $7.99/month | Focused on content generation, not the target-writing document itself |
| School-provided SEN management systems | Whatever your school or trust has adopted for SEND record-keeping | Usually included in school licensing | Not all systems include AI drafting features yet |
General Assistants for Target Drafting
A general chatbot is well suited to turning a paragraph of informal teacher notes into properly structured target language quickly, especially when you specify the Code of Practice area of need and ask for a measurable success criterion explicitly. Treat the output as a strong first draft to edit, not a document to copy directly into a pupil's plan.
EduGenius for the Materials That Support a Target
Once a target is set — say, a Year 3 pupil's target around blending CVC words — a content generator becomes useful for producing the actual practice materials that support it. EduGenius can generate differentiated worksheets and flashcards with a class-profile feature that lets you set a pupil's specific ability range, so materials supporting an IEP target come out at an appropriately scaffolded level rather than needing manual adjustment each time.
This matters particularly for SEN Support pupils, where the gap between a whole-class worksheet and what an individual pupil can actually access is often exactly what the target is trying to close. Generating a scaffolded version of the same core activity — smaller phoneme sets, extra visual support, a shorter task length — takes minutes rather than the manual redesign a fully bespoke resource would otherwise require.
A Worked Example: Drafting a Year 4 Communication Target
Walking through a concrete example makes the workflow easier to apply than a general description alone.
Say you're a SENCO working with a Year 4 pupil who has a diagnosed speech and language need, and your informal notes say: "finds it hard to stay on topic in group discussion, often interrupts, understands one-step instructions well but loses track with two-step ones."
- Prompt: "Draft a SMART IEP target for a Year 4 pupil, SEND Code of Practice area: communication and interaction. Current attainment: follows one-step instructions reliably, struggles with two-step instructions and staying on topic in group discussion. Include a measurable success criterion and a termly timeframe."
- Review the draft against what you know. Check whether the suggested success criterion — for example, "will follow a two-step instruction correctly in 4 out of 5 observed attempts by the end of term" — matches a realistic pace of progress for this specific pupil, based on your own observation and any speech and language therapist input.
- Adjust the wording to match your school's usual phrasing. Many schools have a house style for target language; edit the AI draft to fit rather than importing an unfamiliar structure into the pupil's file.
- Add family-facing context if needed. Ask the tool for a plain-language version of the same target to share at the next review meeting, then check the tone before it goes to parents.
This whole process, done carefully, typically takes five to ten minutes rather than the fifteen to twenty a fully manual draft requires — with the time saved going into the writing step, not the judgment step.
Using AI for the "Review" Stage of Assess-Plan-Do-Review
The SEND Code of Practice's assess-plan-do-review cycle expects each target to be revisited on a set schedule, usually termly, and AI can help with the writing burden of that stage just as it can with the initial draft.
Drafting Review Summaries
Once you have evidence of progress — observation notes, a work sample, or assessment data — AI can help turn that evidence into a clearly worded review summary faster than starting from a blank page. The same rule applies as at the drafting stage: AI turns evidence you already have into readable language; it doesn't generate the evidence or decide whether the target was actually met.
Setting the Next Target
If a target has been met, the next step is usually a new, appropriately more ambitious target in the same area, or a shift to a different priority area entirely. AI can draft either direction quickly once you specify which — "the pupil met the previous two-step instruction target; draft a next-step target building on that skill" produces a more useful result than starting the whole process from scratch.
Mistakes to Avoid
A handful of habits separate AI-assisted target drafting that saves real time from drafting that creates more work later.
- Copying AI-generated target language directly into a plan without editing. Even well-worded AI output needs to reflect what you and the family actually know about this specific pupil — generic-sounding targets undermine the whole point of an individualised plan.
- Skipping the measurable success criterion. A target that says "will improve reading" without a specific, observable measure fails the "measurable" part of SMART and is genuinely harder to review honestly at the next meeting.
- Letting AI infer the area of need instead of specifying it. Without naming the Code of Practice category explicitly, AI-generated targets can drift toward generic language that doesn't map cleanly to any of the four areas.
- Using AI-drafted language in family-facing meetings without a tone check. A target that reads fine internally can sound clinical or impersonal to a parent hearing it for the first time; read it aloud before a review meeting.
- Treating a fast draft as a finished decision. The speed AI adds to the writing step can create pressure to move faster on the judgment step too — resist that; the target still needs the same level of professional consideration it always did.
Key Takeaways
- AI is strongest at turning informal teacher notes into properly worded, measurable SMART targets — it should never be the source of what the target actually is.
- Specifying the SEND Code of Practice area of need (communication and interaction; cognition and learning; social, emotional and mental health; sensory and/or physical) produces noticeably better-aligned draft language.
- The EEF's guidance on special educational needs links specific, measurable targets to better pupil outcomes than vague ones — exactly the quality bar AI can help meet faster.
- Every AI-drafted target still needs a teacher's edit against what's actually known about the pupil, not a copy-paste into the plan.
- Social, emotional, and mental health targets deserve the closest human review of the four Code of Practice areas, since missing context is hardest to catch there.
- EduGenius can generate the differentiated practice materials that support a target once it's set, using a saved class profile for the pupil's ability range.
- The time AI saves belongs in the writing step, not the judgment step — resist letting drafting speed compress the review a target actually deserves.
Used this way, AI shifts a SENCO's time from formatting and phrasing toward the part of the job that actually needs a trained professional: deciding what a pupil needs and confirming the target reflects it accurately.
Frequently Asked Questions
Is "IEP" still the correct term to use in English schools?
Not as a statutory term since the 2014 Children and Families Act, which replaced the older system with SEN Support and, for more complex needs, an Education, Health and Care Plan (EHCP). Many schools, though, still use "IEP" informally as the internal working document that sits underneath SEN Support, so the term remains in everyday use even though it isn't the formal statutory label.
Can AI write an entire IEP or SEN Support plan on its own?
No — AI can draft well-worded, measurable target language quickly from a teacher's notes, but deciding what a pupil actually needs requires classroom observation, assessment data, and family input that AI has no access to. Every AI-assisted draft still needs a teacher or SENCO's review and adjustment before it goes into a pupil's actual plan.
How do I make sure an AI-drafted target is actually measurable?
Ask the tool explicitly for a success criterion stated as a number, percentage, or clearly observable behaviour, rather than accepting a vague phrase like "will improve." A genuinely measurable target should let you answer, at the next review, a clear yes-or-no about whether it was met.
Does using AI to draft targets create any data protection concerns?
Treat any pupil-identifying information the same way you would in any other document — avoid entering a child's full name or other identifying details into a general AI tool unless your school has confirmed it's approved for that use, and check your school or trust's data protection policy on AI tools before entering pupil-specific data. Using anonymised or initials-only descriptions when drafting is a reasonable default if you're unsure.
How is an IEP different from an EHCP in England?
An IEP, where schools still use the term, is typically an internal, school-level working document tracking short-term targets under SEN Support, reviewed and updated by the school itself each term. An EHCP is a legal document issued by the local authority for pupils with more complex or long-term needs, involving education, health, and social care input, and it carries statutory force that a school-level IEP document does not.
Related Reading
References
- Department for Education (DfE). (2015). Special Educational Needs and Disability (SEND) Code of Practice: 0 to 25 Years.
- Education Endowment Foundation (EEF). (2024). Special Educational Needs in Mainstream Schools: Guidance Report.
- Children and Families Act 2014.
- Information Commissioner's Office (ICO). (2024). Guidance on AI and Data Protection.
- Department for Education (DfE). (2023-2024). Generative AI in Education: Guidance for Schools and Colleges.