Financial Literacy / Imali

Money skills, in a language you live in.

We are taking what we learnt from isiXhosa maths and applying it to money: plain lessons a young learner can read, hear and practise.

Row of laptops on a classroom desk, each with a pair of headphones resting on its screen.

Nobody is born knowing what interest is. Most of us learnt about money late, and the hard way.

We think a learner should meet those ideas early, in words she already understands.

Why it matters

In a 1Life survey of young South Africans, more than half said they do not know how to build a stable financial future. Fewer than three in ten had a solid monthly budget. More than 45% said they get their financial information from social media.

It is not only a South African problem. In the OECD's PISA 2022 assessment, which tested 15-year-olds in 20 countries and economies, 18% of students in OECD countries did not reach basic proficiency in financial literacy.

How AI is being used to teach money

Around the world, people are trying AI in financial education in a few distinct ways.

  • Practice with consequences. Simulations let a learner make a money decision, see what happens, and try again without losing anything real.
  • A coach that answers back. Chatbots take a learner's own question and explain it at her level, as often as she needs.
  • Lessons made local. Aflatoun, which works on financial education for children, uses AI to help produce localised modules and adaptive lesson plans.
  • Help for the teacher. AI drafts lesson plans, questions and simpler versions of hard texts, which leaves the teacher more time to teach.

Games are doing similar work without AI. Chow Town, built in South Africa by Sea Monster with Nedbank, teaches money and business skills inside a game and has reportedly drawn 1.5 million players.

What is still unproven

Most of this is new. There is a great deal of enthusiasm and not yet much evidence of what it does for a learner's results. In an Australian survey by the children's money app Kit, 69% of parents said AI could improve their children's financial literacy. That is a hope, not a finding.

AI can also be confidently wrong. With money, a wrong answer costs something. So people have to check the work.

How we will approach it

  • Own language first. A learner should be able to read a money lesson in her own language and hear it spoken properly.
  • Checked by people. First-language speakers and teachers check the content before a learner sees it, as they did for the maths voice.
  • Practice, not lectures. Small decisions with visible results, repeated until they make sense.

What comes next

This work is at the start. If you teach, fund or build in this area, we would like to hear from you.

Get in touch

Sources