South Africa is at a critical juncture in its adoption of artificial intelligence, with growing cybersecurity demands and increasing complexity in hybrid and cloud environments. The country's public sector has adopted a cloud-first approach, but this has led to deep dependency on a small number of global providers, resulting in a loss of influence over pricing, terms, and digital trajectory. Experts argue that AI cannot be allowed to follow the same path.
The country's Draft National AI Policy was withdrawn by Minister Solly Malatsi in April 2026 due to fictitious academic citations, highlighting a governance embarrassment. An Independent Expert Review Panel, chaired by Professor Benjamin Rosman, was appointed to rescue the process. Meanwhile, the private sector has moved forward with AI adoption, with South Africa ranking 46th out of 147 economies in Microsoft's Global AI Diffusion Report for Q1 2026.
The concept of sovereign AI is often misunderstood as protectionism, but it actually refers to owning enough of the AI stack to exercise genuine choice. South Africa needs to build its own domestic capability to avoid structural dependency on global providers. The country has made progress in AI research, with the University of Pretoria ranking first in South Africa for AI research output and the University of Cape Town's AI compute initiative expanding access to processing power.
However, infrastructure limitations pose a significant challenge, with only 26% of South African households owning a computer. Generative AI usage among the unemployed has reached over 90% globally, but in South Africa, access to AI often relies on smartphones with intermittent data and unreliable electricity. The IMF estimates that AI could lift sub-Saharan Africa's economy by around 4% over the next decade, but only if the region improves electricity supply, internet access, and digital skills.
The cost of compute is directly linked to energy cost, which is affected by Eskom's ongoing structural instability. Building a sovereign AI economy requires a stable grid that can guarantee uptime. The Siemens Energy example, which partnered with HPE to build an AI-enabled engineering platform, is cited as a model worth borrowing, demonstrating that global partnerships can be beneficial while maintaining control over data and critical systems.
Countries moving fastest on AI development are building ecosystems that make AI an endogenous capability, rather than an imported service. South Africa has the university depth, data assets, and private sector dynamism to compete, but lacks policy coherence to make these pieces work together. The country's strategic infrastructure operators, such as Eskom and Transnet, could pursue a similar approach to Siemens Energy.
The decision about whether South Africa is building its AI future or merely buying it is being made now, and the country cannot afford to confuse these two decisions again. With the intelligence layer of the economy being built, South Africa must prioritize policy coherence and infrastructure development to ensure that it exercises control over its AI future.
Key points
- South Africa needs to balance access to AI with control over its AI future to avoid structural dependency on global providers.
- The country's infrastructure limitations, including a lack of computer ownership and unreliable electricity, pose significant challenges to AI adoption.
- Policy coherence and ecosystem development are crucial for South Africa to build a sovereign AI economy.
