Africa Will Win the AI Race or Sell It. There Is No Third Option.

Somewhere in the Democratic Republic of Congo, a ton of cobalt leaves a mine for a few hundred dollars. By the time it comes back as the battery in a phone, a car, or a data center, it is worth more than a hundred times what was paid for it. Every dollar of that increase was created somewhere else. That single journey is the story of Africa's century in miniature — and the one we still have time to rewrite.

Because Africa holds the raw materials of the next century. The sun that could power the world's data centers. The cobalt, lithium, and rare earths that go into every battery and every chip. The youngest, fastest-growing workforce on earth, at the exact moment the rest of the world is aging. On paper, no continent is better positioned to lead both the energy race and the AI race.

That is the opportunity. It is also the frustration, because the position on paper and the outcome on the ground keep drifting apart.

The advantage is real

Two races will define the next twenty years, and they are the same race. Artificial intelligence runs on power. The countries that can generate cheap, abundant, clean energy will host the compute, and the countries that host the compute will shape the intelligence that runs the global economy.

Africa can supply both sides of that equation. The solar and hydro potential is enormous. The mineral wealth is already in the ground. The talent is already being born. Put those three together, under African control, and you do not have a market to be sold to. You have a continent that sets the terms.

The people building it know the clock is running. MTN's CEO, Ralph Mupita, recently said Africa must be "obsessed and paranoid about not being left behind" in the AI race — and put his finger on the exact bottleneck: AI runs on power, and Africa's energy deficit is the constraint. Today the continent hosts roughly 2 percent of the world's data centers. Two percent. That is not a talent gap or an ambition gap. It is a gap in infrastructure and ownership — which means it is a gap we can close, if we decide to.

The pattern that keeps repeating.

Here is the honest part, and it is worth saying plainly rather than pretending otherwise.

Too often, the resource leaves and the value leaves with it. A mine is signed away for a quick payment and foreign ownership. Raw ore is exported to be refined somewhere else, at ten times the price, by someone else. A power project is built to serve an export contract before it serves the people living beside it. The money arrives fast, the deal looks good in the moment, and the long-term capability — the refining, the manufacturing, the ownership, the compounding wealth — is built on another continent.

And it is no longer just about minerals. The same pattern is now repeating with data. Professor Benjamin Rosman of the Wits Mind Institute warns that without the skills to build AI at home, Africa becomes "little more than exporters of raw digital resources, dependent on foreign companies to refine our data for us" — only to import the finished technology back "at high cost and with poor local fit." It is the cobalt story again, told in code instead of ore. Fewer than 2 percent of Africa's roughly 2,000 languages are even represented in the large language models the world now runs on. A continent is being trained out of its own intelligence layer.

None of this happens because anyone is foolish. It happens because immediate need is real and patient capital is scarce. When a government or a company can take certain money today or bet on uncertain capability over a decade, the pressure to take the money today is enormous. That is not a character flaw. It is a structural trap, and the one that must be named before it can be broken.

Fifty-four nations, one race.

The second obstacle is fragmentation. Africa is not one actor. It is fifty-four nations, each with its own priorities, its own creditors, and its own reasons to cut a separate deal. A continent negotiating as fifty-four can always be divided. A continent that coordinates — on standards, on refining, on where the value is captured, on how its energy and data are governed — negotiates from a completely different position.

This is the uncomfortable truth behind the opportunity: the advantage is continental, but the decisions are national. Leading the power race and the AI race does not only take sunlight and minerals and young people. It takes cooperation between governments that have every short-term reason not to cooperate, and the willingness to trade a fast individual win for a larger shared one.

The scale of what is needed makes coordination non-optional. The ITU estimates Africa needs some $96 billion in digital infrastructure investment by 2030 to reach parity with more developed regions. No single nation raises and deploys that alone, or should. As Mupita put it, whether AI bridges Africa's divide or deepens it "will be determined by the urgency and unity of our response." Urgency and unity — not one or the other. That is the whole challenge in three words.

What actually changes it.

It is easy to be cynical here. If the incentives always favor the quick sale, and wealth is what lets people leave rather than stay and build, then nothing changes,, and the resources keep flowing outward. I understand that view. On some days I have held it.

But cynicism is a forecast, not a fact, and it is a forecast that quietly excuses inaction. The pattern is not a law of nature. It is the product of specific choices about where value gets captured, and those choices can be made differently:

Refine and build at home. Export the battery and the processed metal, not just the ore. Keep the valuable step on the continent.

Own the infrastructure. Structure energy and data projects so the asset, and the capability to run it, stays under national control rather than being handed over with the contract.

Coordinate where it counts. Move together on standards, pricing, and governance, so no single nation is picked off alone.

Reward patient capital. Back the investors and partners who build capability over a decade, not the ones who extract value in a quarter.

Give talent a reason to stay. The most powerful thing wealth can buy is a door out. The answer is not to resent that — it is to build enough opportunity at home that the most capable people no longer have to choose between success and staying.

This is why we build the way we build

I do not write this from the sidelines. It is the reason my group builds the way it does.

Through Sovera Technologies, we build sovereign AI — systems hosted in-country, under the client's control, with data that does not leave the jurisdiction to go and enrich someone else's model. It is the same principle as the cobalt, applied to intelligence: the valuable step, and the ownership of it, stays on the continent instead of being signed away with the contract. Africa does not need to import its intelligence layer as a finished product and rent it back forever. It can own the capability, govern it to its own standards, and build on top of it.

That is the whole argument in one company. Not resources out and dependency in, but capability built and held at home. The energy race and the AI race will be won the same way — by refusing the quick sale and keeping the value where it was created.

The choice, honestly stated, Africa does not lack the ingredients to lead. It has the energy, the minerals, and the people. What the moment asks for is harder than any of those: the patience to build instead of sell, and the unity to negotiate as a continent instead of as fifty-four separate deals.

That is a real challenge, and I will not pretend it is easy or already won. But the alternative is to watch the raw materials of the next century leave the way the last century's did, and to call it inevitable. It is not inevitable. It is a choice — and it is still open.

The power race and the AI race are Africa's to lead. The only question is whether Africa decides to.

Michael Padilla-Pagan Payano is a CEO, founder, and investor. Through Sovera Technologies, his group builds sovereign, governed AI capability across Africa, the Middle East, and frontier markets — technology a nation can own, not rent.