South Africa's Financial Sector Conduct Authority (FSCA) and the Prudential Authority (PA) jointly published their inaugural report, Artificial Intelligence in the South African Financial Sector, providing a first comprehensive overview of machine learning and generative AI use across South Africa’s financial institutions, informed by a 2024 survey and global developments. The report finds a steady increase in AI usage, led by banks where 52% of institutions report actively employing AI, followed by payment providers at 50%. While most institutions planned investments below ZAR 1 million, more than half of bank respondents anticipated investing over ZAR 20 million in AI technologies during 2024. It highlights potential gains in data analytics, operational efficiency and cybersecurity, but also identifies risks including data privacy concerns, bias and discrimination, reputational risks and systemic vulnerabilities, with regulatory uncertainty, skills shortages and challenges around explainability and governance constraining wider adoption. Key learnings include applying recognised explainability methods such as SHAP and LIME, strengthening governance through data governance, model risk management and board oversight, clearer disclosure when AI is used in consumer-impacting decisions such as credit assessments or insurance pricing, coordination with the Information Regulator on alignment with the Protection of Personal Information Act, and envisaged sector-wide guidance for ethical and responsible AI alongside stronger digital and AI literacy.
South Africa Financial Sector Conduct Authority 2025-11-24
South Africa's Financial Sector Conduct Authority and Prudential Authority publish inaugural report on AI adoption and risks in financial institutions
South Africa's Financial Sector Conduct Authority and Prudential Authority released their first report on AI in the financial sector, noting increased adoption in banks and payment providers. The report highlights benefits in data analytics and cybersecurity, alongside risks like data privacy and systemic vulnerabilities, with regulatory uncertainty and skills shortages hindering broader use. Recommendations include enhancing governance, improving AI explainability, and aligning with the Protection of Personal Information Act.