Curriculum

An AI-literacy map you can inspect

A good curriculum should be legible to parents and teachers, and explainable by children. Here are our competencies and assessment by age.

The through-line is: see AI in the world → understand AI through body and data → ask whether AI is trustworthy → make meaningful work with AI → take responsibility to people, community, and the future.

We place a human-centred mindset, AI ethics, AI techniques and applications, and AI system design into one competency system that deepens with age, rather than bolting them on after the features.

Aligned to an international framework

Competencies by age reference the dimensions of an international AI-literacy framework, combined with children's-rights design principles to ensure age-appropriateness, transparency, and inclusion.

Competency structure adapted from UNESCO's AI Competency Framework for Students (2024), licensed CC BY-SA 3.0 IGO; derived content on this page is released under the same license and marked as modified.

Goals & tools by age

Ages 7–9

See & distinguish

  • Tell a 'rule-based' system from one that 'learns from data'
  • Say that AI makes mistakes, and raise observations in a group
  • Do simple sorting and embodied games to understand data
Paper & image cardsScratchJr / OctoStudioTeacher-led Teachable Machine
Ages 10–12

Collect & test

  • Explain rule-based vs. data-driven
  • Begin discussing bias, privacy, and fairness; make a model card
  • Take part in collecting, labelling, and simple image/sound classifying
ScratchMachine Learning for KidsTeachable Machine
Ages 13–15

Model & critique

  • Describe the AI project lifecycle; compare model vs. result quality
  • Run model tests; judge the reliability of content
  • Plan a community-project prototype with design thinking
Scratch / ML for KidsIntro PythonTensorFlow Playground
Ages 16+

Design & take responsibility

  • Treat AI as an object of research, design, and critique
  • Assess performance, bias, feasibility, and social impact together
  • Complete a showcase prototype and an impact brief
Python + JupyterApp Inventor / web prototypeData-viz tools

Three tracks, not one test

We use three streams of evidence together, valuing the thinking process and reflection over feature completion alone.

Observation record

Teamwork, question quality, and debugging attitude, recorded continuously by teachers in class.

Work evidence

Data cards, model cards, prototypes, and display boards showing how work was designed and tested.

Learner self-review

What did I change my mind about, what do I trust less now, and what do I want to test next.