By Roger B. Jantio*
As the Spring Meetings of the World Bank Group and the International Monetary Fund came to a close, it is useful to step back and connect the threads of this three-part reflection on artificial intelligence and Africa.
The first article examined financing, arguing that borrowing—while sometimes necessary—cannot be the starting point for building AI ecosystems. The second turned to international experience, showing that successful models are grounded in value creation, with capital following rather than leading.
This third and final piece turns to Africa itself.
The question is no longer theoretical. It is practical: what is happening on the ground, and what explains the gap between activity and scale?
The starting point is to recognize that Africa is not operating in a policy vacuum. Far from it. A recent mapping by the Carnegie Endowment for International Peace identifies more than 500 digital policy instruments across over 30 African countries, spanning infrastructure, platforms, skills, and innovation. In parallel, the African Union has articulated a continental digital transformation agenda aligned with its long-term development vision.
In other words, the continent has produced frameworks, strategies, and ambitions at scale.
And yet, outcomes remain uneven. Internet penetration, for example, remains below 40 percent across the continent, reflecting persistent constraints in access, affordability, and infrastructure. More importantly, even where connectivity exists, it does not consistently translate into scaled digital or AI-driven systems.
This leads to a more precise framing of the issue: Africa is not starting from zero, but activity alone is not the same as momentum.
This distinction becomes clearer when looking at how different countries are progressing.
In Kenya, the widespread adoption of mobile money created a foundation for digital services that reach millions of users daily. This has enabled the emergence of data-rich environments in which AI applications can be meaningfully deployed, particularly in financial services, agriculture, and logistics. The lesson is not that Kenya has “solved” AI, but that scale at the application layer creates the conditions under which more advanced technologies become viable.
Nigeria presents a different dynamic. It has one of the continent’s most vibrant startup ecosystems, supported by a large domestic market and increasing access to venture capital. Yet this energy remains fragmented. Companies grow within sectors—fintech, e-commerce, logistics—but the absence of integrated systems limits their ability to scale across the broader economy or regionally. The result is innovation that is real, but not cumulative.
Rwanda illustrates a third approach, characterized by strong state coordination. Digital policies, infrastructure investments, and institutional frameworks have been developed in a relatively aligned manner. This has enabled rapid implementation and clarity of direction. However, coordination alone does not guarantee scale. In the absence of large, interconnected markets or deeply embedded private sector ecosystems, the impact remains more contained.
South Africa offers yet another perspective, with deeper industrial and research capabilities. AI is already being deployed in sectors such as finance, mining, and telecommunications. But even here, the challenge lies in diffusion—extending these capabilities beyond established industries into broader segments of the economy.
Taken together, these examples suggest that Africa’s challenge is not the absence of progress in any single dimension. Rather, it is the lack of integration across dimensions.
The Carnegie analysis is helpful in this regard because it categorizes digital policy efforts into four pillars: infrastructure, platforms, skills, and innovation. On paper, this is a coherent architecture. In practice, these pillars often evolve in parallel rather than as parts of a unified system.
This is where the gap lies. It is not a lack of effort, but a failure of alignment.
Infrastructure is being expanded, but often without sufficient alignment with applications that generate sustained usage. Skills are being developed through training programs and academic initiatives, yet absorption into scaled enterprises remains limited. Innovation ecosystems are encouraged, but startups frequently lack the pathways—regulatory, financial, or market-based—to expand beyond initial pilots. Platforms exist in key sectors, but they are rarely designed for interoperability across borders.
Each of these gaps has tangible consequences.
When infrastructure is not linked to applications, utilization remains low, and investment efficiency declines. When skills are not absorbed into productive systems, talent either migrates or remains underutilized. When startups cannot scale, capital cycles remain shallow, limiting reinvestment and ecosystem growth. When platforms are confined within national boundaries, markets remain too small to support the kind of data generation and network effects that AI systems require.
These are not abstract constraints. They directly affect the ability of African economies to move from experimentation to scale.
They also explain why increasing the number of strategies or mobilizing additional capital, on their own, will not resolve the issue. Without integration, more inputs simply produce more fragmented outputs.
The implication is that the next phase of Africa’s AI development must focus less on adding new elements and more on connecting existing ones.
This requires a shift in emphasis.
First, from infrastructure-first approaches to application-driven systems that create immediate and sustained demand. Second, from nationally bounded initiatives to regional frameworks that enable scale across markets. Third, from fragmented funding mechanisms to capital structures—particularly equity—that are aligned with long-term growth and capable of supporting expansion beyond initial stages.
It also requires a different way of thinking about policy itself. Rather than producing standalone strategies for each pillar, policymakers may need to focus on how decisions in one domain affect outcomes in others—how, for example, data governance frameworks interact with platform development, or how skills programs align with the needs of scaling enterprises.
None of this implies that Africa must follow a single model. On the contrary, the diversity of experiences across the continent is a strength. But that diversity must be anchored in systems that are capable of generating and sustaining value over time.
The building blocks are already present.
The question is whether they can be assembled into something greater than the sum of their parts.
As this series of reflections comes to a close, one conclusion stands out. Africa’s AI future will not be determined by the number of strategies it produces or the volume of capital it mobilizes. It will be determined by its ability to connect activity into systems, and systems into scale.
Until then, progress will remain visible—but incomplete.
Sir Roger Jantio is the Senior Managing Director and CEO of Sterling Merchant Finance Ltd and affiliated investment funds, and an early investor in AI and frontier-tech ventures in the United States and beyond. He is also a strategic advisor with over 36 years of experience in capital allocation and cross-border deal structuring across African markets. Roger is a graduate of Harvard Business School. This concludes a three-part series written during the World Bank Spring Meetings
