Africa AI startups are not short on ambition. According to Google’s Managing Director for Africa, Alex Okosi, founders across the continent are already embedding artificial intelligence into core products. The problem is not the ideas. It is the ground beneath them.
Why Africa AI startups are hitting a ceiling
Okosi points to three pressure points. First, infrastructure gaps make reliable cloud computing difficult outside a handful of cities. Second, available cloud capacity on the continent remains limited relative to demand. Third, and perhaps most consequentially, investors from the Global North are not committing capital at the scale the moment requires.
Together, these constraints do not just slow growth. They prevent African startups from competing at the same level as peers in Southeast Asia or Latin America, where infrastructure investment arrived earlier and in larger volumes.
The stakes are significant. Analysts estimate that AI could add $1.5 trillion to Africa’s GDP by 2035. That figure depends, however, on whether the enabling conditions arrive in time. At present, they are lagging behind the founders who need them.
What this means for Kenyan businesses and consultancies
For businesses operating in Kenya, Okosi’s diagnosis carries a practical message. AI adoption is not simply a technology decision. It is also an infrastructure bet. A Nairobi consultancy advising clients on digital transformation must therefore account for connectivity reliability, cloud access costs, and the availability of local technical talent.
Moreover, Kenyan startups building AI tools face the same funding gap Okosi describes. Local angel networks and regional venture funds are active, but they cannot yet substitute for the deeper pools of growth-stage capital that Global North investors typically provide. This means many promising products stall at the prototype or early-traction stage, before they reach the scale needed to prove out unit economics.
The Kenyan government has signaled interest in supporting the tech sector through various policy initiatives. However, policy intent and infrastructure delivery are different timelines. Founders and consultants working in this space need to plan around the gap between the two.
Building anyway
None of this stops the work. Across Nairobi, Kampala, Lagos, and Accra, developers are building AI tools for agriculture, healthcare, finance, and logistics. They are doing so with leaner resources and less forgiving infrastructure than their counterparts elsewhere. That is not a reason for celebration in itself. It is, however, evidence that the demand for AI solutions in Africa is real and founder-driven, not imported.
For any business in Kenya considering AI adoption, the practical takeaway from Okosi’s remarks is straightforward. Start with problems that are genuinely local and well-defined. Choose tools and vendors with realistic infrastructure requirements. And build relationships with partners who understand the specific constraints of operating in East Africa, not just the global AI market.
The infrastructure will improve. Capital will follow demonstrated traction. But the window between now and 2035 is shorter than it looks.