In February 2026, Anthropic signed its first-ever deal with an African government. The agreement was not with Nigeria — population 220 million, the continent’s largest tech market, the country that generates more AI policy conversation than anywhere else on the continent. It was not with South Africa, whose private sector AI ecosystem is the deepest and most internationally connected in sub-Saharan Africa. The deal went to Rwanda.
Anthropic’s partnership with the Rwandan government focuses on malaria control and cervical cancer screening — two of the country’s stated national health priorities — with AI deployed as a diagnostic and delivery tool. Rwanda’s Minister of ICT and Innovation, Paula Ingabire, described the partnership publicly as “an important milestone in Rwanda’s AI journey” that “reflects Rwanda’s broader strategy to integrate AI into core national systems.” That framing — AI as a component of national systems, owned and governed by Rwanda’s own benchmarks — is the key to understanding what Kigali is building and why the rest of the continent should be paying close attention.
Rwanda’s Strategy Is Not Accidental
None of this happened by chance. Rwanda has been quietly building the institutional infrastructure that makes it the obvious partner for global AI companies for nearly a decade, long before AI was a continental policy priority.
In 2016, Rwanda became the first country in the world to deploy Zipline’s drone delivery system at national scale, using drones to deliver blood and essential medicines to remote health centres. Carnegie Mellon University built its first African campus in Kigali. The Kigali Innovation City — a $2 billion smart city project housing four universities, startup incubators, and research facilities — is operational. The Rwanda Centre for the Fourth Industrial Revolution (C4IR), established in partnership with the World Economic Forum, became the continent’s primary institutional anchor for Fourth Industrial Revolution policy discussions. Microsoft ranked Rwanda among the top five African countries for AI readiness in its 2025 Global AI Diffusion Report.
Then, in April 2025, Rwanda hosted the inaugural Global AI Summit on Africa. Organized by C4IR and the Ministry of ICT in collaboration with the World Economic Forum, the summit drew representatives from over ninety-seven countries under the theme “AI and Africa’s Demographic Dividend.” What the summit produced was not merely aspirational language: the Africa Declaration on AI was signed by ministers from forty-nine countries, the Africa AI Council was formally established under Smart Africa’s umbrella, and a $60 billion Africa AI Fund was announced — with Rwanda’s C4IR positioned to play a central coordinating role.
That same week, Rwanda signed an MOU with the Bill and Melinda Gates Foundation to establish the Rwanda AI Scaling Hub, backed by $7.5 million over three years and focused on AI applications in health, agriculture, and education. In January 2026, a separate $50 million commitment from the Gates Foundation and OpenAI targeted AI integration in Rwandan healthcare. Then came Anthropic.
For anyone still treating Rwanda’s AI positioning as a coincidence, the sequence is difficult to explain. Rwanda is hosting the institutions, signing the MOU agreements, and closing the AI lab partnerships with a consistency and speed that reflects deliberate strategy rather than lucky geography.
What Rwanda Has That Bigger Countries Don’t
The infrastructure arguments for Rwanda’s AI leadership position are well-documented. The governance arguments are less frequently examined.
Rwanda’s national AI policy, published in 2022, is one of the most substantive on the continent. Crucially, it is not organized around innovation promotion — the default frame for most African AI strategies — but around governance. The policy includes ethical oversight committees, risk assessment mechanisms, and explicit frameworks for evaluating AI systems against local development benchmarks. Paula Ingabire’s public statements consistently emphasize that Rwanda’s AI partnerships will be “measured against Rwanda’s own development benchmarks” — a stance that distinguishes the country from AI partnership models elsewhere that accept whichever deployment frame a global technology company offers.
That governance credibility matters to AI companies because it reduces partnership risk. Anthropic is not deploying into a regulatory vacuum in Kigali. It is deploying into an environment with a published AI policy, a Ministry with a clear mandate, an oversight structure with designated institutional actors, and a government track record of following through on technology commitments. For an AI company navigating its first African government deal, that environment is substantially less complex than the alternative.
Anthropic’s signing of its first African government deal received coverage across the continent’s tech media — and each story underscored the same question: why Rwanda and not Lagos? The honest answer is that Rwanda made it easier to say yes. The deal structure is clearer. The governance environment is more legible. The political will is explicit and institutionally supported.
The Concentration Risk Nobody Wants to Acknowledge
The Rwanda AI hub story carries a tension that its most enthusiastic coverage tends to sidestep. Rwanda is a country of fourteen million people. It has no domestic consumer tech market of meaningful scale. The AI hub it is building is, by construction, not primarily for Rwandans. It is a bid to become the governance and institutional home of continental AI — to be to Africa’s AI future what Singapore is to Southeast Asia’s financial infrastructure.
That is a legitimate and coherent strategy. It may succeed. But its success is not cost-free for the rest of the continent.
When global AI companies choose Rwanda as their African institutional anchor, they are implicitly choosing who governs African AI. The Africa Declaration on AI, signed in Kigali, was shaped primarily by the countries with the capacity and institutional presence to participate in its drafting. Nigeria’s Minister Bosun Tijani was on stage. Paula Ingabire was the host. The countries whose citizens will ultimately live under the frameworks being developed — Ethiopia, Tanzania, the Democratic Republic of Congo, Cameroon — were largely present as signatories rather than architects.
The datafication of African economies has consistently followed a pattern where the countries with the earliest institutional infrastructure set the terms that others inherit. Rwanda is doing in AI what South Africa did in financial services regulation: moving first, building credibility, and becoming the reference point. The $60 billion Africa AI Fund’s initial infrastructure commitment — 12,000 Nvidia GPUs distributed across the Big Four nations and Morocco — leaves thirty-nine countries with no computing infrastructure commitment at all.
The gap between Rwanda’s AI trajectory and the median African country’s AI capacity is not narrowing. It is widening, and Rwanda’s strategy is partly responsible for that widening. A continent where one governance hub absorbs the major partnerships, hosts the summit, and sets the norms is not necessarily a continent where AI develops equitably.
Should the Rest of the Continent Be Worried?
The honest answer depends on what “the rest of the continent” wants AI to do.
If the goal is AI as an economic development tool — deployed quickly, at scale, in health and agriculture applications across fragile states — then Rwanda’s hub model is an accelerant. It provides a trusted, governable institutional anchor for global AI companies that would otherwise find most African markets too complex to navigate. The Anthropic deal is more likely to produce useful AI health applications in Rwanda than a comparable negotiation with a less institutionally mature partner would. That benefit is real.
If the goal is AI sovereignty — African countries owning the governance, the data, the deployment frameworks, and ultimately the economic returns from AI systems trained on African data and deployed to African populations — then the concentration in Kigali is a structural problem. AI governance hubs tend to be self-reinforcing. Institutional capacity, partnership networks, and political credibility compound. The countries that are not in Kigali today are less likely to be shaping the frameworks that emerge in 2028.
Rwanda’s national digital payment infrastructure, eKash, built entirely by African engineers and launched in 2025, is emblematic of the country’s broader approach: deploy credibly, build locally, use the deployment as proof of concept for the next partnership. It is a coherent playbook. It is a playbook designed for Rwanda’s interests, not for the continent’s.
The rest of the continent should not be worried about Rwanda building an AI hub. It should be asking who is building theirs.
