South Africans are increasingly turning to AI for financial guidance, revealing a gap the financial services industry can no longer afford to ignore.
AI may be making financial information easier to access, but it is also introducing new risks that consumers may not recognise. For financial institutions, the challenge is no longer whether consumers will use AI, as they already do. The real question is how to turn that behaviour into better financial decisions without compromising trust, accountability, or consumer protection.
Notably, the 2026 Old Mutual Savings & Investment Monitor (OMSIM) found that 54% of working South Africans do not know where to seek financial guidance. Yet 48% report using general AI chatbots for financial guidance, compared with 40% who consult a financial adviser. This is a clear signal that people are seeking accessible, immediate answers to financial questions that traditional channels have failed to provide.
The OECD has highlighted both the opportunities and risks associated with AI in personal finance, including improved accessibility and personalisation alongside inaccurate outputs, bias, commercial influence and privacy concerns. These concerns are particularly important when financial decisions carry consequences that may take decades to reverse.
Why consumers are turning to AI
Part of AI’s appeal is its accessibility, allowing consumers to ask questions at any time, explore multiple scenarios and request explanations without feeling embarrassed about their financial circumstances. Someone struggling with debt, for example, may be reluctant to disclose poor financial decisions to an adviser. An AI chatbot offers an apparently non-judgemental environment in which to explore those concerns. There is also the novelty of being able to ask almost any question and receive an immediate, seemingly authoritative response. But these advantages introduce risks that are not always obvious.
As consumers explore hypothetical scenarios, AI systems may build conversational context around assumptions that bear little resemblance to their real circumstances. The resulting responses can become increasingly aligned with the scenarios and preferences users introduce.
The danger is that consumers receive advice that reinforces what they want to hear rather than what they need to understand. AI can be remarkably convincing, even when it is incorrect. Confidence of expression is not evidence of accuracy.
Financial information is not financial wisdom
Consider someone contemplating a withdrawal from their retirement savings. An AI tool might produce an appealing calculation that fails to account adequately for tax consequences, lost investment growth or the individual’s wider financial circumstances. An experienced financial adviser, however, offers professional judgement, an understanding of individual behaviour and accountability for the advice provided. There are also established channels for seeking advice, raising complaints and seeking recourse. A general-purpose chatbot does not offer equivalent protections.
Interestingly, OMSIM found that only 23% of respondents trust AI to make financial decisions. Consumers appear to recognise the difference between exploring financial questions and relying on technology to make consequential decisions.
The opportunity is therefore not necessarily to replace advisers, but to help consumers become better informed before engaging with advisers. AI can help people understand terminology, compare scenarios and formulate more useful questions. Used appropriately, it can strengthen the relationship between consumers and financial professionals.
Innovation must begin with the problem
Financial institutions face a similar challenge in deploying AI responsibly. The Financial Sector Conduct Authority’s inaugural research into AI adoption found that banks lead the financial services industry in adoption, with 52% of surveyed banks reporting they use some form of AI. Payment providers follow closely, with 50% adoption. Retirement funds reported 14% adoption, while insurers and lenders showed the lowest levels, at 8% each. However, it is acknowledged that AI is not a single solution with a universal measure of success.
Using technology to detect potential fraud, automate administrative processes, or help employees retrieve information is fundamentally different from allowing AI to interact directly with customers about their financial decisions. Each application entails different risks, responsibilities and expectations.
A chatbot that reduces servicing costs but leaves customers frustrated because they cannot reach a human adviser is not necessarily an innovation success. The correct starting point is not where AI can be introduced, but which customer problem needs solving.
At Old Mutual, this thinking informs our approach to experimentation and innovation. Our Moneyversity financial education platform, which has attracted nearly a million users, demonstrates how familiar digital experiences can make financial learning more accessible. We have also developed Capital Gambit, a chess-based game that introduces financial concepts through decisions involving saving, investment, risk and wealth preservation.
These initiatives reflect a broader conviction: meaningful innovation often arises from applying technology and established ideas to enhance people’s capabilities. We are also exploring financial simulations that help consumers understand how different decisions could affect their futures, as well as opportunities to equip advisers with better tools and information.
Trust must determine what comes next
AI presents a particularly important opportunity for Africa. Mobile technology has already helped extend access to financial services beyond traditional banking infrastructure. The next challenge is to ensure that this access translates into financial understanding and better decisions.
AI could help deliver personalised financial education at scale, across languages and income groups, while reducing barriers that have historically limited participation. However, this potential depends on sound architecture, rigorous security testing, appropriate governance, and a clear understanding of where human intervention remains essential.
The industry must also recognise that the rapid availability of AI has outpaced our collective understanding of how best to use it. We are still discovering the innovations that will emerge from the underlying technology.
For financial institutions, the responsibility is to experiment carefully, test assumptions, and scale only those applications that demonstrate genuine customer value. This includes strengthening advisers’ capabilities rather than assuming that automation should replace human judgement.
As Africa navigates accelerating technological change like AI, business leaders must translate innovation into practical solutions that support economic growth, financial inclusion, and long-term resilience. This will be a central consideration at the SingularityU South Africa Summit, taking place from 20 to 22 October 2026, under the theme Igniting Africa’s Future. With Old Mutual as headline sponsor, the Summit brings together business leaders, innovators and investors to explore how emerging technologies and new approaches to leadership can address some of the continent’s most pressing challenges. Importantly, the conversation must extend beyond the potential of technology to how it can be deployed responsibly, at scale, to deliver measurable benefits for businesses and society.
Ultimately, the fact that South Africans are seeking financial advice from AI tells us something important about unmet needs, accessibility and shifting expectations. It does not tell us that consumers no longer need professional advice, accountability or protection.
Trust is becoming the defining currency of AI in financial services. The institutions that earn it will not necessarily be those deploying the most technology, but those using it to help people make better decisions, with greater confidence and appropriate safeguards. That is the measure of innovation that matters.
