South Africa should treat sovereign artificial intelligence as a strategic economic priority and build greater domestic control over the infrastructure, data, skills and systems underpinning AI, according to President Ntuli, Managing Director of HPE South Africa.
As artificial intelligence becomes embedded across businesses, public institutions and critical infrastructure, Ntuli argues that South Africa faces an increasingly important choice: develop sufficient local AI capability to maintain control and choice, or become heavily dependent on technologies designed, trained and governed elsewhere.
The issue extends beyond where AI models are hosted. It concerns who develops the technology, where data is processed, who controls critical computing infrastructure, which rules govern its use and how much of the economic value generated by AI remains within South Africa.
“AI is not simply another technology layer or productivity tool,” Ntuli says. “It is fast becoming the operational intelligence layer of organisations, governments and economies.”
His argument comes as South Africa is simultaneously attracting new digital infrastructure investment, expanding domestic AI research capacity and reconsidering its national AI policy framework [1][2][3].
Why Sovereign AI Is Becoming an Economic Issue
The concept of sovereign AI has become increasingly important as governments and businesses assess their dependence on global cloud platforms, foundation models and computing infrastructure.
Sovereignty in this context does not necessarily mean technological isolation or attempting to recreate every component of the global AI ecosystem domestically. Instead, it involves maintaining sufficient control over strategically important data, infrastructure, intellectual property and decision-making capabilities.
Ntuli draws a parallel with South Africa’s earlier transition towards cloud computing. Cloud adoption accelerated digital transformation, but reliance on a relatively small group of international providers also demonstrated how technology adoption can create long-term dependencies.
“We cannot afford to repeat that mistake with AI,” he argues.
The concern is particularly relevant because AI is moving beyond standalone applications. It is increasingly being integrated into financial services, industrial operations, telecommunications, healthcare, energy, logistics and government services.
South African President Cyril Ramaphosa has also framed digital sovereignty as an increasingly important national issue. Speaking at the Google Cloud Summit in July 2026, he said sovereignty in the digital age increasingly depends on a country’s ability to secure its data, develop domestic digital capabilities and exercise meaningful control over technologies on which its economy depends [1].
South Africa Is Building Domestic AI Compute Capacity
South Africa is not starting from zero.
One significant development is the University of Cape Town’s African Compute Initiative (ACI), announced in March 2026. The initiative is intended to establish what UCT describes as Africa’s largest GPU-intensive computing cluster dedicated to AI research at a higher education institution [2].
The infrastructure will combine modern graphics processing units, multi-petabyte secure storage and high-speed networking, giving African researchers greater ability to develop, train, fine-tune and test AI systems locally.
The initiative forms part of the AI for Development programme, a £58 million partnership co-funded by the United Kingdom’s Foreign, Commonwealth and Development Office and Canada’s International Development Research Centre [2].
UCT says the cluster is expected to support approximately 100 active users during its first year and aims to reach 300 users across at least five institutions by the end of its third year [2].
For Ntuli, investments of this kind are important because sovereign AI depends not only on data centres but also on people capable of developing and adapting AI systems.
“Sovereign AI is not just about where AI runs, but also who builds it,” he says.
That requires researchers, developers, entrepreneurs and organisations to have access to computing resources that have historically been expensive and difficult to obtain.
Digital Infrastructure Investment Is Increasing
Private-sector investment is also expanding South Africa’s digital infrastructure base.
In April 2026, Cassava Technologies pledged R3.6 billion in investment in South Africa through Liquid Intelligent Technologies, Cassava Intelligence South Africa and Africa Data Centres [3].
The investment is planned across a 24-month period and is intended to expand the group’s infrastructure and technology capabilities in the country.
Such investments matter because advanced AI requires significantly more than software. Competitive AI ecosystems depend on data centres, high-performance computing, reliable electricity, high-capacity networks, cloud infrastructure and specialised technical skills.
Energy availability is particularly important. Large-scale AI workloads can require substantial computing power, making the availability and cost of electricity an increasingly important factor in determining where AI infrastructure is located.
Infrastructure Could Determine Africa’s AI Dividend
The economic potential is considerable, but it is far from guaranteed.
An International Monetary Fund analysis published in 2026 estimated that wider AI adoption could raise Sub-Saharan Africa’s GDP by around 4% cumulatively over the coming decade under stronger adoption scenarios [4].
However, the IMF identified electricity supply, digital infrastructure, internet connectivity, technical skills and institutional capacity as major constraints on the region’s ability to capture those benefits.
The Fund estimated that productivity gains could range from only around 0.2% under limited adoption to approximately 2.1% under broader adoption scenarios. Extending AI use into sectors including agriculture could significantly increase the potential economic impact [4].
This makes infrastructure policy part of AI policy.
Countries that cannot provide reliable energy, affordable connectivity, sufficient compute capacity and advanced skills may remain consumers of AI services without capturing a comparable share of the investment, intellectual property and productivity gains generated by the technology.
South Africa Is Reworking Its National AI Policy
South Africa’s policy environment is also evolving.
Cabinet initially approved a Draft South Africa Artificial Intelligence Policy for public consultation in March 2026. The proposed framework included six pillars covering talent development, inclusive growth and job creation, responsible governance, ethical and inclusive AI, cultural preservation and international integration, and human-centred deployment [5].
However, the draft was subsequently withdrawn after problems were identified with references contained in the document. Government said the policy would be reworked to ensure that it establishes credible national standards for the ethical use of artificial intelligence [6].
Communications and Digital Technologies Minister Solly Malatsi subsequently announced an Independent Expert Review Panel to assist with reviewing the policy [7].
The episode illustrates a wider challenge. South Africa needs a regulatory framework capable of encouraging investment and innovation while providing credible rules governing AI development and deployment.
For businesses considering major investments in computing infrastructure, models and AI products, regulatory certainty can become an important component of long-term investment decisions.
Building for Choice Rather Than Isolation
Ntuli stresses that sovereign AI should not mean separating South Africa from global technology ecosystems.
The enormous capital requirements associated with frontier AI make it unrealistic for most countries to compete directly with the largest global model developers and hyperscale cloud providers across every part of the technology stack.
A more practical strategy is to identify which capabilities South Africa needs to own, control or strongly influence.
That could include strategically important datasets, local computing infrastructure, cybersecurity capabilities, AI skills, regulatory frameworks and systems used in sensitive industries or critical infrastructure.
International technology can still play an important role within such a model. The distinction is between participating in global technology ecosystems from a position of choice and becoming structurally dependent on external systems.
What Siemens Energy’s AI Infrastructure Shows
Ntuli points to Siemens Energy as an example of how an organisation can combine advanced global technology with greater control over strategically important systems.
In June 2026, HPE announced that Siemens Energy was deploying a dedicated global high-performance computing platform through GreenLake for engineering workloads [8].
The infrastructure, deployed across sites in the United States and Germany, is designed to integrate accelerated computing and AI into a sovereign private-cloud environment while allowing Siemens Energy to retain control over its systems and data.
The platform supports applications including advanced engineering simulations, digital twins and predictive maintenance. According to HPE, some accelerated simulation workflows are expected to reduce processes that previously took days or weeks to hours [8].
The example demonstrates that AI sovereignty does not necessarily require organisations to develop every technology themselves. Instead, infrastructure can be designed so that businesses retain greater operational control over strategically important data and workloads while still working with global technology partners.
Could the Model Apply to South Africa’s Critical Infrastructure?
Ntuli believes similar thinking could eventually be relevant to strategic South African infrastructure operators, including Eskom.
Energy systems are becoming increasingly data-intensive, while AI can potentially support demand forecasting, predictive maintenance, infrastructure monitoring and complex engineering simulations.
For critical infrastructure operators, however, adoption also raises questions about cybersecurity, resilience, data governance and dependence on external technology providers.
This is where the sovereign AI debate moves beyond the technology industry itself.
If AI increasingly influences the operation of energy systems, financial infrastructure, telecommunications networks and government services, decisions about who controls the underlying infrastructure become questions of economic resilience as well as technological efficiency.
From AI Consumer to AI Builder
South Africa therefore faces two related challenges: accelerating AI adoption while simultaneously building enough domestic capability to avoid becoming only a consumer of technologies developed elsewhere.
The country already possesses several components required for a stronger AI ecosystem: established universities and research institutions, a developing technology sector, growing data-centre investment, financial and telecommunications industries capable of adopting advanced technologies, and an expanding startup ecosystem.
What remains is connecting these assets through infrastructure investment, skills development, credible policy and access to capital.
“The goal is not isolation from global technology ecosystems,” Ntuli says. “It is ensuring that we are building enough of a foundation that we create choice, not dependency.”
That distinction may ultimately determine how much value South Africa captures from the AI transition.
The global AI infrastructure build-out is already underway. Models are being trained, data centres constructed and long-term technology relationships established. Decisions made during this period could shape where intellectual property, technical skills, investment and economic value accumulate for years to come.
For South Africa, sovereign AI is therefore becoming less a debate about technological self-sufficiency and more a question of strategic economic capability: determining which parts of the AI economy the country needs to control, which it can share with global partners, and where dependence could create unacceptable long-term risks.
Sources and Information
[1] South African Government — President Cyril Ramaphosa: Google Cloud Summit.
[2] University of Cape Town — UCT to Lead Africa’s First Higher Education Dedicated AI Compute Initiative.
[3] Cassava Technologies — Cassava Technologies Pledges R3.6 Billion Investment in South Africa.
[4] International Monetary Fund — Unlocking the Potential: AI in Sub-Saharan Africa.
[5] South African Government — Cabinet Statement: Draft South Africa Artificial Intelligence Policy.
[6] The Presidency — Cabinet Statement: Withdrawal of Draft Artificial Intelligence Policy.
[7] Department of Communications and Digital Technologies — Budget Vote Speech — Minister Solly Malatsi.
[8] HPE — Siemens Energy Chooses HPE to Transform Engineering with AI.
This article is based on commentary by President Ntuli, Managing Director of HPE South Africa, supplemented with publicly available information and independent sources.