A recent global study on how people use ChatGPT reports three findings schools can act on now: (1) education is already a major use case; (2) writing and editing dominate workplace use; and (3) the greatest value comes from decision support. Youth adoption is high and growing, with strong implications for equitable AI literacy.
- Education use is significant: tutoring/teaching is a notable share of global usage.
- Writing dominates work tasks: especially editing, critiquing, and summarising.
- AI as co-pilot: strongest value is decision support for informed judgement.
- Equity: youth adoption is high; growth in low- and middle-income regions.
1) Education is already a major use case
Large-scale usage data shows that tutoring and teaching comprise a meaningful portion of global ChatGPT interactions. This validates what we see in classrooms: students increasingly use AI as a study companion, and teachers use it for planning and feedback.
- Embed responsible AI literacy across KS3–KS5.
- Model effective prompts and critique techniques.
- Use AI for retrieval practice, scaffolding, and explainer variants.
- Publish a clear AI use policy (teaching & learning focus).
- Deliver CPD on "AI as study coach" and academic integrity.
- Audit access to ensure equity for all learners.
2) Writing and editing dominate workplace usage
Globally, the largest share of work-related interactions are writing tasks: emails, reports, feedback, and documentation. Importantly, most usage involves modifying user text rather than generating from scratch—an opportunity to teach iterative drafting and ownership of ideas.
- Peer-review with AI: students ask for clarity/style suggestions, then justify accepts/rejects.
- Versioning: show drafts → AI-assisted edits → student rationale.
- Rubric alignment: train AI to critique against your marking criteria.
- Faster feedback with tone/style controls.
- Smarter reports (summaries, differentiation notes, next steps).
- Admin lift: agenda writing, minutes, parent comms templates.
3) AI's biggest value: decision support
Nearly half of global usage is "Asking": seeking advice to make better decisions. In schools, this translates to AI as a thinking partner—testing options, surfacing risks, and generating criteria for choice.
- Define the problem → ask AI to list decision criteria.
- Generate options → prompt for pros/cons with evidence.
- Stress test → ask for failure modes and mitigations.
- Decide → document rationale and next actions.
4) Coding use is smaller—but targeted
Programming queries are a minority of total usage. For CS/Digital Technology, the highest-value patterns are explanation, debugging, and test-case generation—not wholesale code generation.
- "Explain this function" and "predict the output" prompts.
- Generate unit tests, then fix failing cases.
- Refactor for readability and complexity reduction.
- Require design notes & rationale, not just code.
- Oral defence of solutions for authenticity.
- Logging: students record AI prompts/decisions as an appendix.
5) Equity and access
Usage is growing fastest among young people and in low- and middle-income regions. Schools should ensure AI supports—rather than widens—the digital divide.
- Guaranteed access points (libraries, supervised labs).
- Clear guidelines for assistive use (SEND inclusion).
- Transparent citation norms for AI-assisted work.
Implementation roadmap (quick wins)
- Publish a 1-page AI classroom guide for staff/students.
- Run a CPD on "AI for feedback & planning".
- Pilot an AI-assisted writing workflow in one year group.
- Embed AI literacy outcomes across KS3–KS5.
- Adopt decision-support flows in SLT processes.
- Monitor equity and publish an access plan.
Author note: This reflection synthesises insights from recent global usage research on ChatGPT and translates them into actionable steps for schools and teachers.
