Utvecklingssamtal often becomes the moment where everything converges: learning evidence, wellbeing, next steps, and family communication. When time is tight, the risk is either vague language (“keep trying”) or overly teacher-led talk that reduces pupil agency. A practical alternative is to treat the process like an operating system: a repeatable workflow that takes a small set of teacher inputs and generates consistent outputs you can refine. If you already use a brief for parent meetings, you’ll recognise the same logic in this one-page conversation workflow—only here we extend it into an LGR22-aligned package.
In LGR22, the utvecklingssamtal is not simply a “progress chat”. It is a structured conversation that should make learning visible, clarify what the pupil can do, and agree what happens next. The emphasis on pupil participation is not decorative; it shapes how you phrase questions, how you select evidence, and how you write follow-up actions.
The biggest practical shift is between Years 1–5 and Year 6+. In Years 1–5, the IUP (individual development plan) is typically the core written outcome: concrete targets, strategies, responsibilities, and review points, expressed in accessible language. From Year 6 upwards, pupils and families increasingly need a bridge into grading-criteria language. That does not mean the conversation becomes “about grades”, but it does mean you must connect evidence to criteria and show what “next” looks like in the same terms pupils will meet in assessment. If you want a wider view of how LGR22 intentions translate into workable tools, the mapping in LGR22 three years on: gap-to-tool map is a useful companion.
An AI-supported workflow is only worth adopting if it is safe by default. The simplest principle is a “minimum-data” rule: the AI should see only what it needs to produce the draft, and never what it does not.
In practice, that means you can use anonymised learning evidence and teacher observations, but you must never paste sensitive personal data, full health details, diagnoses, detailed safeguarding information, or identifiable incident narratives. Avoid uploading screenshots from school systems. If you would not comfortably read it aloud in a staffroom, do not put it into an AI tool.
Anonymising can be quick. Replace names with “Pupil A”, remove unique identifiers (rare hobbies, small-community references), and generalise dates (“this term” rather than “12 February”). Keep the evidence specific to learning: “writes a clear opening sentence but loses cohesion after paragraph two” is safer and more useful than “often distracted since parents separated”. For the compliance lens schools are increasingly asked about, this EU AI Act explainer for Swedish schools helps you align practice with emerging expectations.
To keep the workflow repeatable, treat your inputs as three “large blocks” you can paste into a prompt.
First, an evidence snapshot: a short, structured view of recent work, assessment notes, and participation. Imagine you are selecting three pieces of evidence you could show the pupil and family without embarrassment or confusion. Second, strengths and needs: what the pupil does well, and where they need support, phrased as learning behaviours and skills rather than fixed traits. Third, next-step priorities: a small set of high-leverage moves for the next period, ideally two to four, that you can actually monitor.
This is the “three inputs in, five outputs out” engine. If you are building wider staff routines, the micro-routine approach in this INSET day AI workshop plan can help you standardise without turning it into a tech project.
The first output is an eight-section, student-led script. The aim is not to make pupils perform; it is to give them a supportive structure that keeps agency and clarity. A Year 4 pupil might read parts aloud; a Year 8 pupil might use it as a speaking frame. Either way, the script should include specific evidence (“Here is a piece I’m proud of…”) and honest next steps (“This is what I’m working on…”), while keeping the teacher’s role as facilitator.
A reliable eight-part structure is: welcome and purpose; what I’m proud of; what I find hard; evidence from my work; how I learn best; goals for the next period; support I want from school and home; and a closing agreement. The AI’s job is to draft phrasing that sounds like a pupil, not a policy document. Your job is to sanity-check: does this sound like the pupil, and does it avoid labels that could stick?
For Years 1–5, the IUP output should be immediately usable: targets, strategies, responsibilities, and review points. The most common failure mode is targets that are either too broad (“improve reading”) or too teacher-centric (“teacher will…”) with little pupil ownership.
Ask the AI to produce two or three targets only, each with a “what it looks like” description and one or two classroom strategies (for example, sentence stems, worked examples, reading fluency routines). Then add responsibilities: what the pupil will do, what the teacher will do, and what home can do, keeping home actions realistic and non-punitive. Finally, include a review point that is time-bound and evidence-based (“in four weeks we will look at two writing samples and a reading record”).
From Year 6, families often want help translating classroom evidence into criteria language without turning the meeting into a negotiation. The “bridge” output works best when it has three parts: criteria language, evidence examples, and “what to do next”.
The AI can draft a short paragraph using the appropriate criteria phrasing (kept general enough to fit your subject and local interpretation), followed by bullet-light examples from the evidence snapshot (“In the last assignment, you…”) and then two concrete next actions. The key is that the bridge should be explanatory, not predictive: it should clarify what quality looks like and how the pupil can move towards it. If your school is also navigating transitions and progression, you may find it helpful to cross-reference how criteria language shifts in this GY25–LGR22 bridge guide.
Ready to Revolutionise Your Teaching Experience?
Discover the power of Automated Education by joining out community of educators who are reclaiming their time whilst enriching their classrooms. With our intuitive platform, you can automate administrative tasks, personalise student learning, and engage with your class like never before.
Don’t let administrative tasks overshadow your passion for teaching. Sign up today and transform your educational environment with Automated Education.
The follow-up letter is where clarity and trust are won. It should summarise what was agreed, what will happen next, and how home can support—without adding new surprises. Build it with a {{name}} placeholder so you can personalise safely after the draft is complete. Keep the tone respectful and practical, and avoid jargon that families may not share.
For Arabic translation, treat the AI as a first-draft translator, not the final authority. Ask for Modern Standard Arabic unless your context requires otherwise, and instruct it to keep names and dates unchanged, preserve bullet/numbering structure, and avoid adding religious or cultural assumptions. Then do a fidelity check: does the Arabic version match the meaning and commitments of the original, including any “we will review on…” details? If you have multilingual support needs more broadly, these minimum viable AI workflows for modersmål can help you build consistent routines.
Finally, produce age-banded student summaries: one for a younger pupil (around Year 2) and one for an older pupil (around Year 8). The content can be similar, but the tone, sentence length, and responsibility framing must change.
A Year 2 summary should be short, concrete, and encouraging, with one clear goal and one clear “how”. It might read like: “You are getting better at hearing the sounds in words. Next we will practise reading a short text every day. You will point to the words, and I will help you when it is tricky.” A Year 8 summary can hold more complexity and agency: it can refer to strategies, evidence, and self-monitoring, while still being supportive rather than managerial. In both, ask the AI to include an “agency line” (“I will…”) and a “support line” (“My teacher will…”), so responsibility is shared.
Before anything leaves your desk, run a short quality check. Accuracy comes first: does every claim match your evidence snapshot, and have any details been invented? Tone comes next: is it respectful, specific, and free from labels that could harm self-concept? Then bias: does the language subtly lower expectations for certain backgrounds, languages, or behaviours? Translation fidelity is its own step: check meaning, dates, and commitments, and ensure the Arabic does not intensify or soften expectations. Finally, record teacher sign-off: a simple note of what was edited and why can be invaluable for consistency and accountability, especially if you are also building audit trails like those described in this report-writing pipeline and data protection audit guide.
A prompt pack only works if it is copy-and-adapt. Keep each template short, and paste your three inputs underneath. Here are compact templates you can reuse.
Use this “system” line at the top of any prompt: “You are drafting for an LGR22-aligned utvecklingssamtal. Do not invent facts. Use only the evidence provided. Keep pupil agency central.”
For the student script: “Create an eight-section, student-led utvecklingssamtal script. Write in first person. Keep it warm, specific, and age-appropriate for Year __. Include two questions the pupil can ask.”
For the IUP (Years 1–5): “Draft an IUP-ready plan with 2–3 targets. For each: what success looks like, strategies in class, what the pupil will do, what the teacher will do, what home can do, and a review point.”
For the Year 6+ bridge: “Draft a grading-criteria bridge: criteria language (general), evidence examples from the snapshot, and two next steps. Avoid predicting grades.”
For the parent letter: “Write a follow-up letter to parents/carers with {{name}} placeholder. Summarise strengths, agreed goals, support actions, and review date. Then provide an Arabic translation in Modern Standard Arabic, preserving structure and meaning.”
For student summaries: “Create two summaries of the same meeting: one for Year 2 (very simple) and one for Year 8 (more detailed). Keep agency, respect, and clarity. Add a final ‘check’ line asking the pupil if it feels fair.”
End-to-end in 20 minutes is realistic once you have the routine: five minutes to assemble the three inputs, ten minutes to generate and edit the five outputs, and five minutes for the quality check and sign-off note. The point is not to automate relationships; it is to protect time for the human parts of the conversation.
May your next utvecklingssamtal feel calmer, clearer, and more pupil-led.
The Automated Education Team
Black Friday can feel like a rare chance to “save” on AI subscriptions, but …
Three years after ChatGPT’s release, schools have enough experience to …
December in schools brings a familiar spike: cover changes, heightened …
Microsoft Ignite can feel like a firehose of AI updates, but schools need a …
Report writing in 2025 is less about “which chatbot is best” and more about …
Three years into LGR22, many schools report real gains in clarity and …
Anti-Bullying Week works best when it moves beyond awareness and into …
Remembrance teaching asks for careful language, accurate sources, and …
Mock season often fails for predictable reasons: revision plans are …
Utvecklingssamtal võib tunduda nagu tosin paralleelset ülesannet: tõendite kogumine, õpilase hääle …
Utvecklingssamtal kan kännas som ett dussin parallella uppgifter: samla underlag, hålla elevens röst …
Utvecklingssamtal voi tuntua tusinalta rinnakkaiselta tehtävältä: näytön kokoamiselta, oppilaan …