CBC Added Digital Literacy. But Who Is Teaching Kids to Question the Algorithm?

Kenyan children are learning to use digital tools. Far fewer are being taught to judge whether what a tool tells them is true. What parents can do about the gap.

The Competency-Based Curriculum did something the old system never managed: it put digital literacy in front of every child as a core competency rather than an optional extra. Children learn to type, to browse, to use educational applications, to find things out.

The gap is what comes after finding out. Kenya now has the highest rate of ChatGPT use of any country in the world, at 42.1% of internet users aged 16 and over in the past month, according to the DataReportal and Meltwater Global Digital Report 2025. Children are using these tools whether or not a scheme of work mentions them, and using a tool is not the same skill as judging what it hands back.

Digital literacy is operational. Can you use the device, find the information, produce the document, stay safe from obvious harm. CBC covers this reasonably well.

AI literacy is evaluative. Do you understand that the tool predicts likely text rather than knowing facts. Can you tell a sourced claim from a confident invention. Do you notice when an answer is fluent but empty. Can you tell whether you are extending your own thinking or replacing it.

The second is harder to teach and harder to assess, which is part of why it lags. It is also the one that decides whether a child is served by these tools or quietly steered by them.

Research on children and AI keeps returning the same shape of finding: most children have been told when they are allowed to use AI, considerably fewer have been taught to use it safely, and fewer still have been taught how to judge whether an answer is accurate.

In other words, permission is being handled. Judgement is not. That is not a criticism of teachers, who are being asked to teach something that arrived faster than any curriculum cycle can absorb. But it does mean the gap is currently a household problem.

Understanding roughly how it works. Not the mathematics. Just enough to know the tool is predicting what text usually follows, which is why it can produce a confident, well-formed, entirely invented answer without any signal that it has done so.

Checking against something else. The habit of verifying one claim in a different source before repeating it. This is an old skill from the library era that transfers directly.

Noticing the difference between expanding and outsourcing. A child who asks the tool to argue against their position is expanding. A child who asks it what their position should be has outsourced.

Recognising an empty answer. AI is very good at producing text that sounds complete and says nothing. Being able to notice this is a genuine skill and children find it satisfying once they can.

Catch it being wrong, together. Ask it something you know well, ideally something local where its knowledge is thinner. Kenyan history, a place you know, a family detail. Let your child watch it be confidently incorrect. That single experience does more than a lecture.

Ask for the thinking, not the answer. When your child shows you homework, ask what they thought before they looked. If nothing, that is the thing to work on, and it has nothing to do with technology.

Ask the school what it teaches, not just what it permits. Most schools can tell you their rule. Fewer can tell you where a child is taught to evaluate an AI answer. It is a fair question and asking it politely helps.

Our Explorers stage, for ages 11 to 12, was built for exactly this gap. Children learn what AI is doing when it answers, how to test it, and how to hold their own position against a confident machine.

It sits between Sprouts, where children build creative confidence before AI appears at all, and Builders, where they use it to make real things. The sequence is deliberate: question it before you build with it.

We also work directly with schools on this, through educator readiness audits and teacher cohorts, because a parent solving it at home is a patch and a department solving it is a fix.

CBC promised competency, creativity and real-world skills. Parents searching for CBC curriculum gaps in Kenya are finding a different reality in many schools: new labels on old methods, portfolio folders, and limited capacity for the creative, technology-integrated learning CBC describes.

Digital literacy was added to the curriculum. But who is teaching children to question the algorithm rather than just click it?

The pattern repeats in three places. Schools teach tool use; children need tool judgment. Schools assess outputs; children need to defend process. Schools follow schemes of work; children need open-ended creation.

Supplementing CBC does not mean rejecting school. It means giving your child environments where competency is practised, not just documented.

We asked ten Nairobi school principals one question: what is your school's policy on students using generative AI for assignments? Not one could answer definitively. Some said they banned it. Some said they had not discussed it. None had communicated anything to parents.

This matches national data: 83% of parents say their child's school has not communicated anything about AI use or policies.

The conclusion is uncomfortable but freeing. You cannot wait for school policy, so set family policy now. Ask your child's teacher directly and create the conversation schools have avoided. Then supplement with structured programmes where AI use is intentional and visible.

If you are the teacher or school leader reading this rather than the parent, AkiliNest runs AI training for educator teams, including the policy and governance sessions that answer exactly the question those ten principals could not.

For families, the kids bootcamps cover the same thinking from the child's side.