For decades, the conversation about inequality has centred on income, employment, and access to capital. Those issues remain fundamental, but another divide is becoming increasingly important: the gap between people who can navigate an increasingly complex financial world and those who cannot.
The internet has put extraordinary amounts of financial information within reach of ordinary consumers, and artificial intelligence is taking this a step further.
It is making sophisticated financial information available instantly, conversationally and at almost no cost. But access to information is different from knowing how to use it, and that distinction could become one of the defining financial challenges of the coming decade.
The World Bank uses the concept of “financial capability” to capture something broader than financial literacy. It encompasses not only what people know, but also the attitudes, skills and behaviours required to manage resources and make appropriate financial choices. That distinction matters enormously in South Africa.
A person can recognise that saving is important yet still be unable to build an emergency fund. They can recognise that expensive debt should be avoided yet still use credit to buy groceries. They can know that retirement provision matters while repeatedly drawing on long-term savings to deal with financial shocks.
Increasingly, the divide is therefore not simply between people who have financial information and those who do not. It is between those who can translate information into decisions and action, and those who struggle to do so.
South Africa's advice gap makes this particularly important
The South African experience offers a powerful illustration. The 2026 Old Mutual Savings & Investment Monitor found that 54% of working South Africans do not know where to seek financial guidance.
At the same time, 48% report using general AI chatbots for financial guidance, compared with 40% who report using a financial adviser. That is an extraordinary shift. It suggests that technology is filling a vacuum that the traditional financial system has not fully addressed.
The important figure may be the 54%. For many consumers, perceptions of the cost and relevance of traditional financial services remain barriers. The financial questions facing an ordinary household may not begin with portfolio construction. They may begin with how to pay the electricity bill, manage several debit orders, or put enough money aside for a child's education.
AI is attractive precisely because there is almost no barrier to asking a question. Yet greater access does not necessarily close the underlying knowledge gap. OMSIM found that only 23% of respondents said they trust AI for financial decisions. This suggests that consumers recognise a distinction between using technology to research a financial question and trusting it to make consequential decisions about their lives.
That distinction should be encouraged.
We have entered an age of abundant financial information
For the first time, a consumer can ask a machine highly personal questions about investing, taxation, debt, budgeting, or retirement and receive a seemingly appropriate answer within seconds. That is an extraordinary opportunity for financial inclusion. It can help people understand terminology, explore options, and enter financial conversations better informed. But there is a fundamental limitation: advice requires context.
The right financial decision depends on far more than the question. It can depend on someone's debt, dependants, income stability, existing investments, insurance, tax position, risk appetite, age, and long-term objectives. An AI system knows only the context a user gives it. The danger is that an answer can sound authoritative even when key pieces of that context are missing. This is why international policymakers are paying increasing attention to the issue.
The OECD's recent work on artificial intelligence and personal finance highlights both the opportunities offered by greater accessibility and personalisation and the risks, including bias, inaccurate outputs, commercial influence, privacy concerns, and financial exclusion.
The IMF has similarly warned that generative AI can produce convincing yet incorrect information, with potentially serious consequences when these technologies are used in financial services.
The challenge of the AI age may therefore be less about accessing financial information than about learning how to interrogate it.
The cost of poor decisions compounds too
We normally talk about compounding in positive terms: invest consistently, earn returns on those returns, and time does much of the work. But poor financial decisions can compound too.
Consider what happens when a household lacks sufficient protection against a financial shock. A burst geyser or an unexpected expense becomes debt. Servicing that debt reduces the capacity to save. Savings may then be withdrawn to meet another expense, weakening the household's ability to absorb the next shock. Eventually, financial resilience is steadily eroded.
OMSIM offers a striking glimpse into this pressure. Some 46% of working people report transferring money between accounts to avoid or manage debit orders. This is not conventional financial management but more akin to financial triage.
Moving money to delay a debit order may provide temporary breathing space, but it can also trigger bank charges, missed-payment fees, and further financial pressure. A short-term coping mechanism can therefore become an additional cost.
The South African Reserve Bank's data provides the broader context. Household debt stood at almost 62% of disposable income at the end of 2025.
Financial education cannot solve an income problem, and we should be careful not to pretend it can. Millions of households face genuine cost-of-living constraints. But knowledge and financial capability can affect how effectively people manage scarce resources.
Saving is not enough if savings never get the chance to grow
This becomes particularly important when we consider how people save. South Africans continue to show a strong desire to put money aside. OMSIM found that the proportion of income allocated to savings among working South Africans remained at 22%. Yet the way money is held matters.
The proportion keeping cash unbanked has risen sharply, from 40% in 2020 to 64% today. There are understandable reasons for this. Cash is accessible, familiar, and immediately available when something goes wrong. For people who have lost money on poor investments, it may also feel safer. But accessibility has a trade-off.
Money repeatedly withdrawn to cover emergencies never has the chance to compound. Cash that earns little or no return can also lose purchasing power over time.
As a result, people may save without necessarily building wealth.
This is where financial capability becomes far more practical than simply teaching people that they "should save". The better question is: what is the money for, when will it be needed, what financial shocks need to be protected against, and where should different pools of savings be held?
The next divide
South Africa's financial knowledge gap should therefore not be viewed merely as a consumer education problem. It has much wider implications.
Two households earning the same income can reach very different financial outcomes over a decade because of the decisions made along the way: how debt is managed, whether savings are preserved, whether money is invested productively, how financial shocks are absorbed, and whether appropriate guidance is sought when important decisions arise.
This is also why AI should not be cast as the enemy of financial advice. Used properly, it could become an extraordinarily powerful financial education tool. It can lower barriers to asking questions and help people build the confidence to engage with their finances.
But information, education and advice serve different functions, and the opportunity is to connect them.
Consumers should be able to use technology to learn, research and ask better questions, while recognising when a decision requires personalised, accountable advice that takes their full circumstances into account.
The most consequential wealth gap of the future may not simply separate those who earn more from those who earn less. It may increasingly separate those who know how to make their money work from those who never had the knowledge, confidence, or guidance to give it a chance.
Sources:
1. https://www.oecd.org/en/topics/sub-issues/financial-education.html?utm_source=chatgpt.com
2. https://www.oecd.org/en/publications/artificial-intelligence-and-personal-finance_2858fdf4-en.html
Share