What Is AI Literacy for Kids? The New International Framework, Explained for Kenyan Parents

The OECD and European Commission have set out what children should be able to do with AI. The four areas in plain language, what they look like at 8 and at 16, and what to ask any AI class for kids.

Most conversations about children and AI in Nairobi are stuck on one question: should my child be using it or not? That question matters, but it skips the more useful one. If children are going to grow up with AI around them, and they are, what exactly should they be able to do with it?

In June 2026 the OECD and the European Commission published an answer. The AI Literacy Framework, developed with support from Code.org and shaped by feedback from more than 2,000 teachers, researchers, policymakers and others in over 100 countries, sets out what learners in primary and secondary school should know, be able to do and value when it comes to AI. It is the closest thing there now is to an international standard, and it is worth a parent's ten minutes.

A child who can get a chatbot to write a composition is not AI literate, any more than a child who can open a book is a reader. The framework treats AI literacy as a mix of knowledge, skills and attitudes: understanding roughly how these systems work, being able to use them well, and having the judgement to question them.

That last part is the one parents tend to feel is missing. Kenya has the highest rate of ChatGPT use of any country in the world: 42.1% of internet users aged 16 and over used it in the past month, according to the DataReportal and Meltwater Global Digital Report 2025. Use is not the gap. Judgement is.

Engage with AI. Understanding what AI is doing when you use it. Noticing where it shapes what you see, describing how it works without pretending it thinks like a person, and deciding whether an answer should be accepted, fixed or thrown away. This area also covers the harder questions: how AI can repeat unfair patterns, and what it costs in energy and resources.

Create with AI. Using AI to take your own ideas further rather than replacing them. Starting from an original idea, trying different tools to prototype it, asking AI for feedback and deciding what to do with that feedback, and being honest about credit and copyright when AI played a part.

Manage AI. Deciding when AI belongs in a task at all. Some work should be done by a person, some can be supported by AI, and a child who can tell the difference, and stay in charge once a tool is involved, is well ahead of most adults.

Shape AI. Understanding that AI systems are designed by people, for particular users, trained on particular data, with limits. Older learners go on to test systems against clear criteria and think about how they could be improved for the communities they serve.

The framework describes each competence at a basic, an intermediate and an advanced level, so the same idea grows with the child.

For a younger child, managing AI might mean sorting everyday school tasks into three groups: ones a person should do, ones AI could help with, and ones AI could do alone. For a teenager, it becomes breaking a research project into steps and deciding, deliberately, which steps need their own voice.

For a younger child, creating with AI might mean brainstorming alone first, then looking at what an AI suggests and comparing the two. For a teenager, it means marking up a piece of their own writing to show which changes came from AI and which were theirs, and explaining why they accepted or rejected each one.

Notice what those examples have in common. The child's own thinking comes first every time, and the tool is something they judge rather than something they obey.

Does my child think before the tool opens? If sessions start with a prompt, the tool is doing the thinking.

Will my child learn when not to use AI? A class that only teaches how to use it covers one of the four areas.

Will my child catch AI being wrong? Checking and rejecting AI output is a skill, and it has to be practised on purpose.

Does the programme talk about bias, data and credit? These sound advanced, but the framework starts them early, at a level a child can manage.

Can my child explain what they made? A finished project a child cannot explain is the tool's work, not theirs.

For more on choosing a programme, see how to choose an AI bootcamp for kids in Nairobi.

We have mapped our four stages, from Sprouts at 8 to Innovators at 17, against the framework's four areas, and added topics where the mapping showed a gap: giving credit for ideas, where AI's data comes from and what it costs the planet, and copyright and bias for older learners. The mapping is our own and has not been reviewed by the OECD or the European Commission, and we show it in full so you can judge it yourself.

See the mapping on the AI for kids programme page, read the framework itself at ailiteracyframework.org, or get in touch to ask us how it plays out for your child's age.