AI in South Africa’s Financial Sector: Key Insights, Risks, and Adoption Trends

TL;DR:

This report by the South African Reserve Bank and the Financial Sector Conduct Authority examines the integration of artificial intelligence within the nation’s financial services. It details how local institutions use machine learning and generative AI to enhance operational efficiency, improve fraud detection, and provide personalized customer experiences. While banking and payment providers lead in adoption, the document identifies significant risks regarding data privacy, cybersecurity, and algorithmic bias. To address these concerns, the authorities emphasize the need for robust governance frameworks, human oversight, and regulatory alignment with international standards. Ultimately, the report serves as a roadmap for fostering an ethical AI ecosystem that supports financial stability and consumer protection in South Africa.

Based on the 2024 joint survey conducted by the Financial Sector Conduct Authority (FSCA) and the Prudential Authority (PA), the following data and statistics outline the state of Artificial Intelligence (AI) adoption, investment, and governance in the South African financial sector.

AI Adoption and Sector Distribution

The survey received approximately 2,100 responses, with 220 respondents (10.6%) currently using AI in the South African financial sector. Adoption varies significantly across segments:

• Banking Institutions: Lead adoption at approximately 52%.

• Payments Institutions: Show 50% adoption.

• Pension Funds: 14% adoption.

• Investment Providers: 11% adoption.

• Insurance and Lending: Both show the lowest integration at 8%.

Intended Investment (2024 Forecast)

Investment intentions reveal a divide between large banking institutions and other financial entities:

• Banking Sector: Over 50% of banks planned to invest more than R20 million in AI during 2024. Specifically, 45% of banks that adopted AI planned to invest more than R30 million.

• Broad Industry Trend: Most institutions plan to invest less than R1 million, indicating a focus on incremental implementation.

• Sector Comparisons: Approximately 62% of investment providers and 41% of insurers intended to spend less than R1 million in 2024.

Primary Applications and Use Cases

Financial institutions distinguish between Traditional AI (machine learning/rule-based) and Generative AI (GenAI):

• Traditional AI Application Areas: Operations and IT (18%) is the primary area, followed by Risk and Compliance (16%) and Sales and Marketing (16%).

• GenAI Focus: Primary applications are found in Sales and Marketing.

• Specific Use Cases: Fraud detection and product/service promotion are the leading use cases for machine learning and GenAI, respectively.

Perceived Risks and Constraints

Despite the benefits, several regulatory and non-regulatory factors limit AI adoption:

• Top Risks: Cybersecurity (37%), Reputational Risk (25%), and Hallucinations/ML Ethics (24%) are cited as pressing concerns.

• Regulatory Constraints: 82% of respondents identified data protection and privacy laws (such as POPIA) as the most significant regulatory challenge.

• Non-Regulatory Constraints:

→ Insufficient talent/access to skills: 53%.

→ Transparency and explainability issues: 41%.

→ Lack of sufficient data: 39%.

Governance and Ethics

Governance frameworks are becoming foundational to AI deployment in South Africa:

• Governance Adoption: 68% of institutions use a Risk Management framework, and 55% use Data Governance.

• Explainability Methods: Feature Importance is the leading method used to understand AI decisions. However, 21% of institutions currently use no explainability method at all.

• Ethical Priorities: Data Privacy is the top ethical consideration at 90%, followed by Ethical Use of AI (80%) and Security (78%).

Macro-Context Statistics

• South African Financial Sector: Contributes approximately 20% to the national GDP.

• Sector Composition: Includes 67 banks, 158 insurance institutions, and over 200 fintechs.

Citation

This summary is based on Artificial Intelligence in the South African Financial Sector, published jointly by the South African Reserve Bank (SARB) and the Financial Sector Conduct Authority (FSCA) in 2025. The report examines the adoption and use of artificial intelligence across South Africa’s financial services industry, with a focus on banking, payments, and other regulated financial institutions. © 2025 South African Reserve Bank and Financial Sector Conduct Authority. All rights reserved. Source: Artificial Intelligence in the South African Financial Sector (PDF).