AI Reshapes South Africa’s Financial Sector
Artificial Intelligence (AI) is no longer a futuristic concept; it has become a fundamental component of South Africa’s financial sector. From customer interactions to fraud detection systems and internal operations, AI is integrated into various platforms including chatbots, call centers, and multifunctional “super apps” on mobile devices.
The Challenge of Governance and Regulation
As AI adoption accelerates, a significant message emerged from the recent panel discussion at the Financial Sector Conduct Authority’s (FSCA) 2026 conference: governance, skills, and regulations are struggling to keep pace with technological advancements. According to Ayanda Ngcebetsha, director of data and AI at Microsoft South Africa, “AI is indeed here. We can no longer wish it away.”
From Pilot Projects to Real Decisions
The panel highlighted that AI in financial services has moved beyond mere experimentation. Darren Franks, co-founder of the FinTech Association of South Africa, shared that fintech firms have an average AI maturity score of 3.45 out of 5. While 14% have live use cases, and 32% are scaling, none have achieved full deployment. Concerningly, 20% of firms lack AI governance, and only 5% have mature oversight frameworks, despite 86% believing AI will be critical for their business in the next five years.
Franks emphasized, “This is not sci-fi. AI is here, and it’s being used in businesses today.” The application of AI ranges from customer service chatbots to fraud detection and compliance monitoring, increasingly influencing customer outcomes.
The Evolving Role of AI
Fatos Koc, head of the financial markets unit at the Organisation for Economic Co-operation and Development (OECD), pointed out that AI in finance is shifting from decision support to decision delegation. “Models are no longer just advising humans. They are executing decisions autonomously at scale,” she noted.
Case Study: Discovery Bank
Discovery Bank’s CEO, Hylton Kallner, provided insights on how the bank utilizes AI to:
- Build a behavioral fingerprint for every client based on spending habits, payment patterns, and geolocation.
- Monitor transactions in real time against this fingerprint.
- Detect potential fraud by analyzing network effects among clients.
This system can issue real-time alerts for suspicious transactions, delay payments in high-risk cases, or even lock down the app if a client is suspected to be under duress. Kallner emphasized the importance of minimizing false positives to avoid disrupting legitimate transactions.
Governance Lags Behind
Despite rapid adoption, oversight mechanisms have not kept pace. Many organizations rely on inconsistent or ad hoc AI governance policies. Ngcebetsha summarized the situation: “Innovation is forging ahead, and policy is catching up.”
The Black Box Problem
Explainability remains a pressing concern. Nolwazi Hlophe from the FSCA identified a gap between technical performance and real-world accountability, stating that a model might be accurate yet still lack transparency in its decision-making process.
She cautioned that insufficient explainability could lead to bias, unfair outcomes, and reputational harm, particularly in sensitive areas like credit decisions and fraud detection. As one audience member articulated, “If you can’t explain it, you can’t audit it properly.”
Amplifying Risks
The panel identified several risks that intensify as AI adoption deepens, including:
- Data privacy and protection
- Cybersecurity vulnerabilities
- Data quality and representativeness
- Third-party dependencies
- Model hallucinations and errors
These interconnected risks could rapidly escalate into systemic challenges, particularly with the increasing reliance on large, general-purpose AI models.
Impact on Jobs
The discussion also addressed the impact of AI on employment. The consensus was that AI is more likely to reshape jobs rather than eliminate them, provided organizations manage the transition effectively. Hlophe noted the importance of upskilling employees to prevent job losses due to AI implementation.
Franks highlighted observable market shifts: a 42% decline in demand for software engineers and a 69% increase for commercial roles, indicating a rebalancing of skill requirements.
African Opportunities and Risks
Ambassador Lavina Ramkissoon of the African Union expressed that Africa has a limited opportunity to leverage AI for growth and inclusion. She emphasized that without clear strategies, AI could exacerbate fragmentation and inequality across the continent.
The Regulatory Balancing Act
For regulators, the challenge lies in enabling innovation while managing associated risks. The panel suggested more adaptable regulatory approaches, including:
- Principle-based regulation
- Regulatory sandboxes and experimentation
- Activity-based oversight
“What works is not slowing innovation, but adapting oversight to its pace,” Koc stated. The overarching message from the discussion was clear: AI has already integrated into the core of financial services. The pressing question is whether institutions and regulators can keep up.
