Report writing season is one of the biggest workload spikes in a teacher's year, and it's also the area where AI use needs the most care — because reports are pupil-specific by definition.
The safe approach
Never paste a pupil's name or identifying details into a public AI tool alongside their attainment data. Instead, use AI for the parts that aren't pupil-specific:
- Generating a bank of generic comment starters and sentence structures for different attainment levels, with no names attached
- Varying your phrasing so comments don't all sound identical across a class set
- Turning your own bullet-point notes about "Pupil A" (anonymised) into full sentences, which you then personalise with the real name afterwards
What stays entirely yours
The actual judgement about a specific child's progress, effort, and next steps should always come from you — AI can help you say it more efficiently, but it should never be deciding what to say. Read every AI-assisted comment back and ask: does this actually sound like something I know about this child?
A realistic time saving
Most teachers using this approach report the wording and phrasing side of report writing — not the judgement side — taking noticeably less time, since the anonymised-notes-to-full-sentences step removes the "staring at a blank box" problem that eats so much of an evening. The professional judgement about what to actually say about each child doesn't get any faster, and it shouldn't.
Key takeaways
- Anonymise first — write "Pupil A," then personalise afterwards.
- AI speeds up wording; the judgement about what to say stays with you.
- Never paste real names alongside attainment data into a public AI tool.