SONA 2025, AI and South Africa’s leadership: Why the national AI strategy can’t wait

As South Africa assumes the presidency of the G20 in 2025, the world is watching how it will shape critical global policies, including artificial intelligence (AI) governance. Across the G20, countries like France, Canada, Germany, Japan and India have already finalised comprehensive AI strategies, ensuring AI is embedded in economic growth, digital transformation and workforce development.

On the African continent, Egypt, Mauritius, Rwanda, Benin and Senegal have also made significant strides in AI governance by developing their own national AI strategies. Additionally, the African Union’s Continental Artificial Intelligence Strategy (2024) provides a framework for responsible AI development, aligning ethical AI adoption with Africa’s unique economic and societal challenges.

However, despite leading the G20 and being home to some of Africa’s top AI research institutions, South Africa has not finalised its national AI strategy. The Department of Communications and Digital Technologies (DCDT) drafted an AI policy plan in 2023, but SONA 2025 failed to mention any commitment to its implementation.

While AI was referenced in fraud detection, visa automation and policing, it was not discussed as part of a structured AI policy framework. These were the only mentions in SONA 2025:

• Fraud detection and tax enforcement – AI-powered risk analysis helped SARS prevent R95 billion in fraudulent refunds and recover R20 billion in revenue.
• Visa processing and travel authorisations – AI is streamlining the Electronic Travel Authorisation system, reducing delays and corruption.
• Smart policing and crime prevention – AI is assisting in crime analytics, predictive policing and forensic investigations to tackle high crime rates.

These are useful applications, but without a governing AI policy, they do not support South Africa’s long-term AI growth, workforce readiness, or economic positioning.

South Africa has made progress in AI development through the establishment of AI research hubs, which support several pillars of the draft AI strategy. However, without a finalised policy, these hubs lack the national funding, governance and regulatory frameworks needed to maximise their impact.

Key AI hubs in South Africa

1. University of Johannesburg (UJ) AI Hub (2022): Focuses on manufacturing, fintech, energy and the criminal justice system.
2. Tshwane University of Technology (TUT) AI Hub (2023): Specialises in automotive, farming, food production, health and telecommunications.
3. Central University of Technology AI Hub (2024): Enhancing AI research and applications at a regional level.
4. Defence Artificial Intelligence Research Unit (DAIRU) at the South African Military Academy (2024): Africa’s first military-focused AI hub, exploring AI’s role in defense and national security.

These hubs support multiple pillars of the national AI policy — from AI innovation and start-up support, to workforce transformation. But without policy alignment, they operate in silos, with no clear roadmap for expanding South Africa’s AI-driven economy.

South Africa’s draft AI strategy outlines 12 key pillars for AI governance and innovation—none of which were referenced in SONA 2025.

These pillars include:

• AI in government and the economy – Encouraging AI-driven economic growth.
• Government-as-a-platform approach – Enhancing AI integration in public services.
• AI public skilling and awareness – Expanding AI education and workforce training.
• AI sandboxes for testing – Creating safe AI experimentation spaces.
• AI innovation and start-up grants – Funding South African AI entrepreneurs.
• Dedicated AI legislation and ethical governance – Regulating AI safety and fairness.
• AI for workforce and economic transformation – Preparing industries for AI disruption.
• AI in strategic sectors (health, energy, transport, etc) – AI deployment for infrastructure.
• AI for social good and inclusion – Ensuring AI benefits all South Africans.

By failing to mention these pillars, South Africa’s AI development remains reactive, rather than proactive.

As G20 President, South Africa has an opportunity to:

• Lead global AI governance discussions on human rights, ethics and responsible AI.
• Ensure AI is part of trade and economic policies at the G20 level.
• Align with international AI governance frameworks, including:
• UNESCO’s AI Ethics Recommendations for transparency and fairness.
• OECD AI principles on responsible AI and sustainability.
• Vatican AI guidelines for ethical AI adoption.
• African Union’s AI strategy, supporting AI-driven economic growth across Africa.

Without a national AI strategy, South Africa risks losing influence in these global AI conversations.

The South African government must act NOW!

• Finalise and implement the national AI strategy — policy action, not just discussion.
• Align AI hubs with national policy — ensuring AI research drives economic transformation.
• Invest in AI education and workforce readiness — AI adoption must support economic inclusion and job creation.
• Use the G20 presidency to shape global AI governance — Africa has an opportunity to lead, not follow in AI regulation.

AI is reshaping economies, industries and governance. South Africa must act now — or risk falling behind.

About Nazareen Ebrahim

Nazareen Ebrahim is an accomplished AI ethicist recognised among the 100 Brilliant Women in AI Ethics 2024, with over 20 years’ experience spanning media, technology and communications. She serves as the AI Ethics Lead at Socially Acceptable, sits on the Minara Chamber of Commerce board overseeing its technology portfolio, and leads premier African PR firm Naz Consulting International. A sought-after international speaker and team leader at the Centre for AI Policy Development’s Advanced AI Policy Clinic, she is currently writing her first book, Africa’s 5IR: How Ethics will Lead the Technology Revolution, due in 2025.

First published on Naz Consulting International’s website and speaker profile for Nazareen Ebrahim.