Africa must not only fuel the AI revolution, it must shape it

I recently attended the International Association for Safe and Ethical AI conference in Paris at Unesco. Among the most striking things Yoshua Bengio, one of the “godfathers of AI”, said during his address was: “If you are not at the table, you are on the menu.”

His point was clear: the Global South and the “middle powers” must collaborate urgently to shape the future of artificial intelligence (AI) or have it shaped for them.

Africa has been on that menu before, in 1913, when the Natives Land Act forced Black South Africans off their ancestral land and into the migrant labour system. Men left their families to dig gold and diamonds from the earth; minerals that built Johannesburg, funded an industrial economy and generated extraordinary wealth for a few. Almost none of it reached the communities whose sweat and land made it possible.

More than a century later, a troublingly familiar pattern is taking shape. In the Lualaba Province of the Democratic Republic of Congo (DRC), miners, some of them children, extract cobalt by hand from tunnels that regularly collapse. That cobalt travels through a global supply chain until it reaches the data centres of Silicon Valley and Shenzhen, where it powers the servers training the most advanced artificial intelligence systems on Earth. The same systems that may, within a generation, reshape every economy on the continent.

Across the world’s leading technology centres, governments and corporations are racing to build artificial general intelligence (AGI), advanced AI machines capable of performing most intellectual tasks as well as, or better than, humans. Elon Musk once warned that AGI poses existential risks, characterising its pursuit as “summoning demons”. He has since changed course, lured by the race to build and control the global economy through AGI, confessing: “I resisted AI for too long… living in denial. Now it is game on.”

The global competition has begun in earnest. Yet Africa is largely excluded from the conversation, not because it lacks relevance, but because it is not at the helm of AI production.

The economic stakes are staggering. If AGI systems can eventually perform labour at extremely low cost, the value of human work, both blue and white collar jobs, could decline dramatically, concentrating wealth among those who own the machines and their underlying infrastructure. For a continent where South Africa alone faces unemployment above 30%, and where millions of young Africans are entering labour markets that already cannot absorb them, this is not a theoretical concern. It is an approaching crisis.

Geoffrey Hinton, widely regarded as one of the pioneers of modern AI, warned after leaving Google that AGI could dramatically worsen economic inequality, enriching a small class of technology owners while displacing large sections of the workforce. The Congress of South African Trade Unions has raised similar alarms, insisting that digital technologies must be introduced in ways that protect workers rather than deepen the divide.

But African voices are doing more than sounding warnings. Rachel Adams, a South African researcher in AI governance, has argued that African countries risk becoming passive consumers of technologies designed elsewhere unless they actively shape the rules governing AI.

Research ICT Africa has warned that without regulatory capacity and infrastructure investment, the continent may be reduced to what it has been before: a source of raw materials and markets, while the value is captured elsewhere. Initiatives like the Masakhane research community – a pan-African network of AI scientists building language models for African languages – demonstrate that the continent already has the intellectual capacity to contribute. What it lacks is institutional backing and political will.

South Africa is currently finalising its national AI policy. This is a critical moment. Given the cross-cutting economic implications of artificial intelligence – touching on industrial policy, trade, education, labour, mineral beneficiation and digital infrastructure – this work should not sit with the Department of Communications and Digital Technologies alone.

The Presidency ought to lead a whole-of-government approach that integrates AI policy with the country’s broader economic trajectory and existing industrial strategies. Artificial intelligence is not a communications issue. It is a national strategic priority that will have a generational impact on our economy.

Beyond South Africa, the African Union and Southern African Development Community should convene an AGI preparedness summit, bringing together heads of state, technology researchers, trade unions and civil society to establish a continental position before the rules of the AI age are written without African input.

As the great AGI debate ensues, African governments hold greater leverage than they realise. The AI industry depends on critical minerals found overwhelmingly on this continent: cobalt from the DRC, platinum group metals from South Africa, lithium from Zimbabwe, graphite from Mozambique. Without these materials, the global AI supply chain collapses. This is our greatest power, but only if it is used strategically and deliberately.

The emerging concept of demand-side governance recognises that influence in the AI ecosystem does not belong only to those who build models. It belongs equally to those who control market access, mineral supply chains and procurement. African governments should align with the growing global movement demanding that safety measures keep pace with AI capability, and condition access to these resources on compliance with international AI safety standards, transparency requirements, technology transfer and investment in African research, education and skills.

This would ensure that the technologies built with Africa’s resources contribute to the continent’s development and humanity’s wellbeing broadly, rather than repeating, in digital form, the extraction that has defined its relationship with global industry for centuries.

AGI could become the most transformative technology in human history. It could accelerate scientific discovery, transform healthcare and unlock new economic possibilities. But it could equally deepen inequality, hollow out labour markets and concentrate technological power in ways the world has never seen.

For Africa, the question is not whether AI will reshape the future. It will. The question is whether the continent will help write the rules, or whether, once again, its resources will build someone else’s prosperity while its people are left to manage the consequences.

The African Union ought to convene that summit. The Presidency ought to lead South Africa’s AI strategy. And the world ought to understand that a global technology built on African minerals requires African consent, African participation and African leadership.

Anything less would be extraction by another name. DM

Mbali Hlophe, an ANC MPL, is the chairperson of the Gauteng Portfolio Committee on e-Government, Policy and Research Development.