Learning goals
- Explain, in plain language, what an AI system is and why its output is not automatically true.
- Use a prompt to ask for an explanation, then improve the question with constraints and context.
- Identify privacy, bias, authorship, and verification choices in a real scenario.
- Create a small project that shows responsible human–AI collaboration—or choose not to use AI and explain why.
The sequence follows UNESCO’s progression of Understand → Apply → Create, with human-centred and ethical reflection in every week. Read the framework.
| Week | Focus | Activities | Evidence of learning |
|---|---|---|---|
| 1 · Understand | What is AI? | Sort examples into “rules,” “search,” and “learned pattern.” Discuss autocomplete, recommendations, and image recognition. Make a paper “prediction” game. | Learner explains one capability and one limitation without using the word “magic.” |
| 2 · Apply | Questions and prompts | Ask for two explanations of the same topic at different ages. Compare specificity, assumptions, and missing context. Practise the Pause → Probe → Prove routine. | Learner revises a vague prompt and names one fact that needs checking. |
| 3 · Question | Safety, bias, and authorship | Review fictional chat transcripts: a private-data request, a biased description, and an invented citation. Role-play how to pause, report, and ask a human. | Learner identifies the risk and proposes a safer alternative for each transcript. |
| 4 · Create | Make a responsible project | Choose a mini-project: a source-backed explainer, quiz, story outline, or family AI policy. Document what the learner did, what AI did (if anything), and how claims were verified. | Portfolio entry plus a two-minute explanation of choices and limits. |
Age adjustments
Ages 8–11: use drawing, sorting cards, read-aloud examples, and a parent-operated tool. Avoid accounts and real personal data. Ages 12–14: add source comparison, citations, and a schoolwork policy discussion. Ages 15–17: examine model limitations, consent, copyright, and the trade-offs of different tools.
Assessment rubric
- Explain: distinguishes a prediction from a verified fact.
- Protect: spots information that should stay private.
- Question: asks for context, uncertainty, or a source.
- Create: keeps human responsibility visible and can describe the process.
Offline option: Print the family AI check-in worksheet and use the paper prompts. No account or AI tool is required to complete this plan.