Africa Is Not Ready for the AI Era – And Few Are Willing to Say It

By Kgothatso Meka, Founder, Thiego Capital

At a recent G20 Africa AI Village event, optimism filled the room. Speakers celebrated new artificial intelligence (AI) strategies, the rise from 80 to 2,000 GPUs, and Africa’s youthful population. The message was clear: Africa is not behind in the global AI race.

But optimism can obscure reality. Africa is not ready for the AI era, and few are willing to say it.

This is not cynicism – it is clarity. And clarity is the starting point for any serious strategy.

Measuring Against the Global Frontier

Progress cannot be measured against our own past. It must be benchmarked against nations shaping the AI future.

In the European Union, the AI Act is the world’s first comprehensive AI law. It is championed at the level of the European Commission Presidency and backed by billions of euros through Digital Europe, Horizon Europe, and the Chips Act. It sits at the heart of industrial policy and a geopolitical agenda –not as a side project of an ICT ministry.

In China, President Xi Jinping has framed AI as central to national rejuvenation. The ‘New Generation AI Development Plan’ has woven AI into education, defence, and health – among countless other domains. It is an inter-ministerial project tied to long term global leadership.

The United States links AI to security, competitiveness, and diplomacy. The $280 billion Chips and Science Act is part of an effort to rebuild American semiconductor and advanced computing capability by funding research factories and talent development.

The Gulf states are equally ambitious. After appointing the world’s first Minister of State for AI, the UAE anchored its AI Strategy 2031 inside the Prime Minister’s Office. Saudi Arabia views AI as a key enabler for nearly 70% of its Vision 2030 goals, committing approximately $20 billion to data centers and cloud platforms to drive economic diversification.

In these countries, AI sits on the desks of presidents and prime ministers. In most African nations, it sits inside ICT ministries – far from the center of power and less likely to drive national strategy. That gap matters.

Africa’s Structural Reality

When measured against the global AI race, Africa’s structural deficits are stark:

• Energy: Approximately 600 million Africans still live without access to electricity. Without reliable power, there can be no compute, no data centers, and no digital public services.
• Data Infrastructure: Despite housing 18% of the world's population, Africa accounts for less than 2% of global data center capacity. Most African data is stored offshore where it falls under foreign control, foreign regulation, and foreign commercial interests.
• Connectivity: Only around 40% of Africans are online, with women disproportionately excluded.
• Digitization: AI cannot operate on paper. Yet, many public systems – health records, school administration, police and court files, and property registries – remain undigitized.
• Policy: Only 16 of 54 African countries have AI policies or strategies. Many are consultant-driven PDFs rather than actionable national plans anchored in the Presidency, Finance, or Planning ministries. Without budgets, political will, or private-sector coordination, they remain shelfware rather than strategy.

Even our youth dividend is at risk. With 60% of its population under 25, Africa’s youth profile is often described as a demographic dividend. But in the AI era, automation threatens the entry-level jobs that traditionally absorb young workers. Without a strategic pivot, hundreds of millions of new labour market entrants represent a potential structural risk, not an automatic dividend.

The Economics of Sovereignty: Threats to Africa’s AI Wealth

According to a 2024 GSMA report, AI could add as much as US$2.9 trillion to Africa’s economy by 2030. This is the prize.

Yet, it is critical to understand that economic value does not automatically remain where it is created. It flows to those who control the foundational elements of AI production: data, compute infrastructure, models, and talent. If Africa continues to engage primarily as a consumer of foreign platforms rather than as a producer, much of this projected value will exit the continent through licensing fees, cloud hosting costs, data extraction. 

The risk is clear: Africa could become one of the largest markets for AI without being a meaningful contributor to its creation. That is not inclusive growth; it is structural dependency. 

Data Colonialism: A Silent Threat to Africa’s Sovereignty

AI is not just an economics issue; it is a question of sovereignty and power.

First, the narrative risk. Africa has lived this story before. Our histories were erased or rewritten, our languages weakened, and our cultural memory interrupted. In the AI era, this risk returns in a new form. If African languages, cultures, histories, and ways of knowing are not embedded inside the systems shaping daily life, Africa becomes invisible inside the models billions will rely on.

Second, the economic extraction risk. In the past, Africa’s minerals were exported raw, refined elsewhere, and sold back at a premium. Today, data is the new gold, and the cycle is repeating. Africans consume foreign platforms and supply behavioural, financial, and other data that trains these systems. Processing and value creation happen offshore. As Dr. Mark Nasila notes, “No nation has ever become a superpower by outsourcing the processing of its natural resources.” The same truth applies to data.

Third, the geopolitical risk. When a nation’s core digital functions – banking, identity, logistics, or government services – sit on servers outside its borders, the power to switch them off sits outside the continent. In a world where digital systems are instruments of foreign policy, this is a critical vulnerability.

To break historical patterns, Africa must own its digital resources. Only then can AI drive locally owned prosperit, rather than a new channel for economic vulnerability.

A Drop in the Ocean: Why 2,000 GPUs Are Not Enough for Africa

These risks clarify why the recent applause for 2,000 sovereign GPUs was misplaced. While a step forward, a look at the global landscape reveals the gap. 

Tech giants are operating on a scale that dwarfs entire continents. Meta, for instance, deployed roughly 350,000 H100 GPUs in 2024 alone. Microsoft committed $80 billion in FY2025 to AI infrastructure, including clusters with 200,000 GPUs to power its Azure cloud.

China has built a massive network of 230 AI clusters with a total estimated stock of 400,000 H100-equivalent chips. Meanwhile, India’s national AI mission started with 10,000 GPUs and scaled to 38,000.

Against this, 2,000 GPUs can support only a handful of meaningful AI projects. For a continent of over 1.5 billion people, it is a symbolic step, not a structural shift.

From Risk to Action: Building a Sovereign AI Stack

Africa needs more than policy documents and pilot projects to harness AI’s potential. By investing in sovereign digital infrastructure, accelerating the digitization of public systems, and fostering locally-driven AI innovation, the continent can shift from dependency to self-reliance. The ultimate goal is to develop a “sovereign stack”: a layered architecture that protects African data, empowers African models. 

Regional collaboration through platforms like the African Continental Free Trade Area (AfCFTA) can aggregate resources for data centers and compute clusters, as outlined in its Digital Trade Protocol. Universities and entrepreneurs can lead in developing models that reflect African languages, cultures, and priorities. 

Building this stack is essential – not just as a shield, but the way to capture the $2.9 trillion prize. Without it, value flows offshore; with it, we unlock new industries and ensure AI becomes a driver of inclusive, locally owned prosperity.

However, this must be pursued pragmatically, blending independence with strategic partnerships. Hybrid strategies, such as leasing compute from global providers under strict local governance, can provide quick wins while infrastructure catches up. 

The Six Layers of Sovereignty:

1. Energy Sovereignty: Data centres are energy vampires. You cannot train African models if the grid in Lagos or Johannesburg is collapsing. Sovereignty starts with the power socket.
2. Infrastructure Sovereignty: Secure data centers, fibre networks, and cable landing points so that African data can live and move within the continent.
3. Data Sovereignty: Public systems must be digitized and governed locally. Data is the raw material for AI; without control of it, African nations will always build on someone else’s foundation.
4. Compute Sovereignty: Data without processing power is inert. Africa needs GPU clusters, sovereign cloud environments, and regional compute capacity to turn raw data into intelligence.
5. Model Sovereignty: African languages, histories, cultures, and priorities must be embedded in the models themselves to prevent further cultural erasure.
6. Talent and Institutional Sovereignty: We need regulators, engineers, and researchers capable of designing, deploying, and governing AI systems responsibly.

These layers are interdependent – energy underpins infrastructure, data fuels compute – but progress need not be strictly sequential. Phased hybrids allow leaps, like using partnered clouds for model development while investing in energy. 

A Call for Strategic Honesty

Africa has immense potential, brilliance, and pockets of genuine progress. But the continent is not ready for the AI era. Saying so is not defeat; it is clarity. And clarity is the starting point for strategy.

We must stop measuring progress by the existence of policy drafts. We must stop celebrating constraint as innovation. We must stop telling ourselves that leapfrogging infrastructure gaps is a sustainable strategy.

The world is not waiting. If Africa does not act now, it risks being left out of the AI economy entirely. But if we act with strategic honesty, we can transform this vulnerability into strength. The time for drafting is over; the time for building has begun.