Africa’s AI window is closing

South Africa has a narrow opportunity to establish itself as one of Africa’s artificial intelligence (AI) powerhouses, but success will depend on whether it can build local capability, expand infrastructure and skills, and develop AI that reflects African languages and realities rather than relying on technologies built elsewhere.

Those are among the key findings of “Africa’s AI Imperative: Building Capability, Sovereignty and Scale for the Continent’s Next Decade”, a report commissioned by Naspers and Prosus.

The report argues that Africa’s participation in the AI revolution is inevitable, but its role is not. The continent faces a defining choice: become a producer of AI or remain largely a consumer of technologies developed in the US, Europe and China.

The report comes as governments worldwide race to establish AI strategies while companies invest billions in infrastructure, chips and software. Yet Africa accounts for less than 1% of the global AI economy despite having the world’s youngest population.

Five priorities are identified for the continent in the report:

• building sovereign AI capability;
• expanding computing infrastructure;
• creating African datasets;
• attracting long-term capital; and
• improving co-ordination among governments, industry and academia.

For South Africa, the challenge is particularly significant because it is widely viewed as one of Africa’s regional AI anchors alongside Kenya, Nigeria and Egypt.

The report says South Africa already has several advantages, including sophisticated financial markets, world-class data centres and a growing ecosystem of AI companies. However, these strengths will matter little without co-ordinated execution.

The choices we make around investment, innovation, governance and skills will determine whether we become creators of value in the global AI economy or remain consumers of technologies developed by others.

—  Khusela Sangoni, chairperson of parliament’s portfolio committee on communications and digital technologies

Khusela Sangoni, chairperson of parliament’s portfolio committee on communications and digital technologies, said the country was at one of the defining moments in its economic history.

“Artificial intelligence is no longer a technology of the future. It is rapidly becoming the infrastructure upon which the future itself will be built,” she said. “The choices we make around investment, innovation, governance and skills will determine whether we become creators of value in the global AI economy or remain consumers of technologies developed by others.”

Rather than seeing regulation as an obstacle, Sangoni argued that effective governance should become a competitive advantage.

“The countries that will succeed in the age of AI will not necessarily be those with the fewest rules. They will be those that create the greatest confidence for entrepreneurs, investors and citizens.”

She said regulation could no longer be developed in isolation because AI is evolving faster than traditional policymaking. Waiting for a perfect framework was not an option, she added, and government and industry would need to work together continuously as technologies develop.

One concern is what Sangoni described as a “reasoning governance gap” in which existing laws regulate AI inputs, such as data, and outputs, such as transparency, but pay little attention to how AI systems arrive at decisions.

She said South Africa should avoid creating another cumbersome regulatory bureaucracy and instead strengthen existing institutions — including the Independent Communications Authority of South Africa, the Competition Commission and the Information Regulator — while improving co-ordination between them.

The report argues that governance should enable innovation rather than constrain it and that trust will ultimately determine the pace of adoption across society.

Artificial intelligence is changing how we work, how we learn, how we govern, how we produce, and increasingly how we compete as nations

—  Phuthi Mahanyele-Dabengwa, group executive director of Naspers and Prosus

Phuthi Mahanyele-Dabengwa, group executive director of Naspers and Prosus, said the AI debate should be viewed primarily through an economic lens rather than purely as a technology discussion.

“Artificial intelligence is changing how we work, how we learn, how we govern, how we produce, and increasingly how we compete as nations,” she said.

Every industrial revolution had created winners and losers, she noted.

“Those that recognised the moment early invested in new capabilities, shaped new institutions and ultimately defined the rules of the emerging economy. Those who hesitated found themselves adapting to systems designed elsewhere.”

Mahanyele-Dabengwa said AI adoption would increasingly become a measure of national competitiveness as governments identify the digital economy as a key growth sector. She argued that local companies should not simply import AI tools but help build them.

“South African-founded companies cannot merely be adapting to the AI revolution by adopting solutions developed elsewhere. Local firms must be at the centre of shaping and leading it,” she said.

The report argues that Africa’s greatest long-term opportunity lies in developing AI systems rooted in local languages, cultures and datasets rather than relying exclusively on imported models. It cites companies such as Lelapa AI as examples of locally developed multilingual models capable of understanding African languages often overlooked by global systems.

The continent does not lack vision, talent or strategy. What remains scarce is the co-ordinated execution required to translate intent into infrastructure, products and policy

—  Report

It warns that if Africa does not build and control its own datasets and computing infrastructure, it risks surrendering both economic value and strategic independence.

Infrastructure remains one of the continent’s biggest constraints. Training advanced AI systems requires substantial computing power, reliable electricity and high-speed connectivity, areas where much of Africa still lags.

The report highlights encouraging developments, including AI factories announced by Cassava Technologies, Dell Technologies and Altron, as well as Uganda’s plans for a sovereign AI supercomputer powered by renewable energy. These investments suggest the continent is beginning to close the infrastructure gap, though significantly more capital will be required.

Talent is another major concern. Africa has a globally competitive pool of AI researchers and engineers, but many leave because local companies cannot yet absorb them at scale. While training programmes are expanding, the report argues that creating enough high-value AI jobs will determine whether the continent captures value from the AI revolution or continues exporting its best talent.

The authors also distinguish between commercial AI and developmental AI. Corporate applications that improve efficiency and productivity are advancing rapidly because they generate immediate returns. However, AI applications in health care, education, agriculture and small businesses continue to struggle for funding and scale despite their potential social impact. Bridging that gap will require patient capital and stronger partnerships between governments, business and development institutions.

Ultimately the report argues that Africa’s AI race will not be won through ambition alone. Success will depend on whether governments, companies and universities can work together to turn policy into implementation, develop trusted governance frameworks and build AI systems that solve African problems.

“The continent does not lack vision, talent or strategy,” the report concludes. “What remains scarce is the co-ordinated execution required to translate intent into infrastructure, products and policy.”

TimesLIVE