Johannesburg, South Africa – In April, African Union ministers gathered in Tangier, Morocco, to discuss artificial intelligence at a moment when governments across the continent are racing to develop AI strategies, attract investment and expand digital infrastructure.

Beneath the enthusiasm, however, sits a more fundamental question. As foreign technology companies invest in data centres, cloud services and AI systems across Africa, how much control will African countries ultimately have over the infrastructure on which those technologies depend?

The debate reflects a broader shift in how policymakers are thinking about AI. For years, discussions focused largely on adoption: how governments, businesses and public services could use the technology. Increasingly, attention is turning to ownership, governance and the terms on which AI systems are developed and deployed.

Several governments have framed the issue in those terms. Nigeria, Kenya, Egypt and Ghana have all released national AI strategies in recent years that highlight the need to build local capacity and reduce dependence on foreign technology providers. Ghana’s national strategy, launched in April, describes AI as a “sovereign capability”. Forty-nine countries, along with the African Union, have endorsed the Africa Declaration on Artificial Intelligence, which calls for greater investment in African AI infrastructure, talent and innovation, alongside proposals for coordinated financing mechanisms.

At the same time, translating ambition into policy has not always been straightforward. In South Africa, a draft national AI policy was withdrawn earlier this year after officials identified references that could not be verified and appeared to have been generated by AI tools, highlighting the practical challenges governments face in regulating rapidly evolving technologies.

Global competition, local leverage

The discussion is unfolding as global competition over AI intensifies. Major technology companies, cloud providers and governments are competing for access to data, computing power and new markets. For African countries, that competition may also create space to negotiate.

Priyal Singh, a geopolitical analyst at Signal Risk, told Al Jazeera that the fragmented nature of the global AI industry could strengthen that position.

“African states will indeed be provided with greater room for manoeuvre on AI and data infrastructure, precisely due to how contested and fragmented this industry is amongst global leaders,” he said.

He pointed to regulatory tensions surrounding Starlink’s expansion in parts of Africa as an example of governments becoming more assertive in their dealings with global technology firms.

“Major tech companies will need to bend to local concerns much more often than they would otherwise expect,” Singh said.

The infrastructure gap

Yet leverage in the AI era is not only political. It is also infrastructural.

Africa remains significantly underrepresented in the global digital economy’s physical backbone. Industry estimates suggest the continent accounts for less than one per cent of global data centre capacity, despite being home to roughly 18 per cent of the world’s population. Research by McKinsey has found that Africa’s five largest data centre markets combined have less capacity than France. Across much of the continent, unreliable electricity supply remains a major constraint on expansion.

Those limits help explain why negotiations over data centres and cloud infrastructure have become increasingly sensitive.

Kenya’s contested data centre deal

One of the most closely watched projects has been a proposed $1bn data centre development involving Microsoft and Emirati technology company G42 in Kenya.

The project drew attention after Kenyan President William Ruto highlighted the scale of its energy demands, warning that infrastructure of that size would require substantial additional power generation.

Reports have also pointed to discussions over commercial arrangements and long-term commitments linked to computing capacity. Kenyan officials have maintained that talks around the project remain ongoing.

Whatever the outcome, the episode illustrates the trade-offs governments face: attracting investment in AI infrastructure while weighing energy needs, financing costs and long-term strategic dependence.

What countries gain and what they give up

The question of who builds Africa’s digital future extends beyond Western technology companies.

Sanusha Naidu, a senior research fellow at the Institute for Global Dialogue, told Al Jazeera that debates about diversification are often more complicated than they appear.

“Whether it’s seen as diversifying from Western tech companies or shifting towards Chinese-based companies, I think it’s generally part of the cost-benefit factor,” she said.

For governments, she argued, the key issue is what is returned through these partnerships.

“Whether it’s a US company, a company from Europe, or a Chinese company,” she said, policymakers must weigh the broader developmental impact of such investments.

She compared current AI infrastructure debates with earlier waves of foreign investment.

“What we saw in the 1990s around the textile industry is investment comes in, but there’s a lot of subsidisation by the recipient country. With data centres, it’s much more intense. It’s also how big consumers of water these data centres are, and how that impacts socioeconomic issues within African countries.”

Data, surveillance and sovereignty

Concerns about dependence extend beyond data centres.

Over the past decade, African governments have adopted a growing range of foreign-built digital systems, from cloud computing platforms and digital public services to surveillance and smart city technologies. At the same time, debates over data governance, digital sovereignty and where sensitive information should be stored and processed have become increasingly prominent across the continent.

Similar arguments have been made by supporters of plans to establish an Africa Credit Rating Agency, designed to offer African-led assessments of sovereign creditworthiness rather than relying exclusively on established international ratings agencies.

The missing public

Yet much of the discussion about AI governance remains concentrated among policymakers, regulators and technology companies.

Joseph Asunka, chief executive of Afrobarometer, told Al Jazeera that the debate is still far removed from everyday citizens.

“These negotiations should not just be conducted at the elite level and dumped on citizens,” he said. “If citizens do not trust their government’s actions in this space, it creates a trust gap, which could have negative implications for the adoption of fintech, e-commerce and e-government tools.”

He added that concerns about data protection and digital security are already widespread across African populations, even if AI itself is not yet widely understood.

Beyond dependency

The debate echoes older questions about economic sovereignty that have shaped African politics for decades. Independence-era leaders argued that political freedom meant little without control over economic resources. Today, similar questions are emerging around data, computing power and digital infrastructure.

Alongside large-scale investment, governments and development agencies are also exploring ways to build local capacity. Projects such as the United Nations Development Programme’s timbuktoo initiative aim to strengthen African technology ecosystems through support for innovation, entrepreneurship and digital infrastructure.

Such efforts remain modest compared with the scale of global AI investment. But they reflect a broader attempt to ensure African countries participate not only as consumers of AI systems, but also as contributors to their development.

Africa is unlikely to become self-sufficient in artificial intelligence, nor is that the objective for most governments. The continent remains deeply integrated into global technology supply chains and will continue to rely on international investment, expertise and partnerships.

The question that remains

The question facing policymakers is therefore less about whether Africa will use AI than about the terms on which it should do so. As governments negotiate new investments, draft regulations and build digital infrastructure, decisions made now could shape who controls the technologies that increasingly influence economies, public services and everyday life.

“These negotiations should not just be conducted at the elite level and dumped on citizens,” Afrobarometer’s Asunka said.

“If citizens do not trust their government’s actions in this space, it creates a trust gap, which could have negative implications for the adoption of fintech, e-commerce and e-government tools.”