By Roger Jantio
The central point of this article is straightforward: Africa’s AI future will not be built by founders alone. Founders are essential. They discover problems. They build products. They test markets. They take risk. They force new possibilities into existence. But even the best founder cannot, company by company, solve unreliable power, weak connectivity, fragmented data systems, unclear regulation, slow procurement, limited compute access, shallow risk capital and cross-border market fragmentation.
At some point, the founder’s burden becomes an institutional question. That does not mean institutions should replace entrepreneurs. They should not. Africa does not need bureaucracies pretending to be startups. It does not need public agencies trying to pick every AI winner.
It needs institutions that understand their real role: to create the conditions under which serious AI companies, infrastructure platforms and data ecosystems can emerge, scale and attract capital. This distinction matters.
In the early stages of the AI conversation, awareness was useful. Conferences, national strategies, public debates and pilot projects helped place AI on the African agenda. But awareness has now done much of its work. The next question is whether Africa can build AI ecosystems that are investable, trusted and scalable.
That requires seven institution pushes.
The first institutional role is trust
AI depends on data. Data depends on trust. Trust depends on institutions. In health, finance, education, agriculture, insurance and public services, the issue is not simply whether data exists. It is whether the data is usable, reliable, lawful, protected and trusted enough to support serious AI systems.
A hospital may have records, but are they digitized? A government may have registries, but are they accurate? A bank may have transaction histories, but can they be responsibly used? A ministry may hold valuable public data, but under what rules can it be accessed? A local AI company may need data to train, fine-tune or validate a model, but who owns the data and who has the right to use it?
These are not technical questions alone. They are institutional questions. Africa should not treat data sovereignty as a slogan. It should treat data as an economic asset that must be governed, protected, licensed and negotiated. This requires credible rules on privacy, consent, data access, cross-border transfer, cybersecurity, public-interest use and benefit-sharing.
Without such rules, two bad outcomes become likely. Either data remains locked away in the name of sovereignty, limiting innovation, or it is extracted by external platforms that capture most of the value elsewhere. Neither outcome serves Africa well.
The better path is governed data value. Institutions can help create that path by building trusted data frameworks, secure data environments, sectoral data-sharing arrangements, public registries, standards for anonymization, and procurement rules that protect citizens while allowing responsible innovation.
This is especially urgent in health. Bad data in health is not merely inefficient. It can be dangerous. Consent, clinical sensitivity, data quality and institutional accountability are not theoretical issues. If African health AI is to become investable, patients, clinicians, regulators, hospitals, insurers, governments and entrepreneurs must be able to trust the data and the rules around it.
The second institutional role is infrastructure
AI is often discussed as software. But serious AI capacity rests on physical and digital foundations: electricity, connectivity, compute, cloud access, data centers, cybersecurity, cooling, edge devices and digital public infrastructure.
A founder can design around some constraints. A strong company may build offline-first tools, operate in low-bandwidth environments, rely on implementation partners or reduce compute costs through better architecture. But no founder should be expected to rebuild the operating environment alone.
This is where institutions matter. Governments, development finance institutions, infrastructure investors, telecom operators, utilities, data-center developers, cloud providers, universities and regional bodies all have a role to play. They can help lower the cost of building, reduce execution risk and make it easier for companies to move from isolated deployments to repeatable scale.
This does not mean every African country must build the same AI infrastructure. That would be wasteful and unrealistic. Some countries may become sophisticated adopters. Others may host regional data centers or compute capacity. Some may build strengths around digital public infrastructure. Others may specialize in fintech, health, agriculture, logistics, energy, education or AI-enabled services.
Africa should not try to build one abstract “African AI ecosystem.” It should build a network of specialized nodes.
This is how technology ecosystems often develop. In the United States, Silicon Valley, Boston, New York and Seattle did not become strong in the same way. Silicon Valley built around software, platforms and venture capital. Boston built around life sciences and research institutions. New York built around finance, media and enterprise markets. Seattle built around cloud, commerce and engineering depth.
Africa will need its own pattern. Lagos, Nairobi, Kigali, Johannesburg, Cape Town, Cairo, Tunis, Accra, Casablanca, Dakar and other centers will not all play the same role. Each should build from strengths already in the ground: fintech, mobile infrastructure, governance systems, energy, financial services, logistics, research, language capabilities, public digital systems, trade corridors or entrepreneurial density. The institutional task is not to force uniformity. It is to enable specialization, interoperability and scale.
The third institutional role is market access
Many African AI founders can prove a first use case. The harder question is how they move beyond one pilot, one customer, one city or one institution. That transition often depends on procurement channels, sector partnerships and trusted buyers.
This is where institutions can either accelerate or block innovation. If public procurement is slow, opaque or hostile to young companies, AI startups will struggle to sell into government even when their products solve real public problems. If hospitals, banks, insurers, agribusinesses and utilities are unwilling to partner with startups, founders will struggle to access real workflows and real data. If regulatory pathways are unclear, investors will hesitate. If public agencies prefer large foreign vendors by default, local AI companies may remain subcontractors in their own markets.
Institutions must therefore become smarter buyers and better partners. Governments do not need to buy every AI product. But they should create procurement pathways that allow qualified local and regional companies to compete. They should use sandboxes where appropriate. They should create clear rules for pilots, data access, evaluation and scaling. They should avoid trapping founders in endless demonstrations that never become contracts.
For AI companies, a pilot that cannot become a contract is not a market. It is a distraction. Institutions can change that by linking pilots to budgets, performance metrics, procurement pathways and scale decisions. This would help serious founders and protect public resources at the same time.
The fourth institutional role is capital formation
Africa’s AI future will require different kinds of capital. Not every opportunity is a venture-capital opportunity. Not every AI company is suited for grants. Not every infrastructure project can be financed with equity alone. Not every public-interest platform can survive without concessional support or guarantees.
Institutions matter because they can help organize the capital stack. Venture capital can support high-growth AI startups. Development finance institutions can help de-risk markets, support governance, mobilize private capital and back commercially credible companies with development impact. Strategic investors can provide distribution, sector access and enterprise customers. Banks can support working capital where revenue is predictable. Donors can fund public-good data, research, capacity building and early experimentation. Governments can anchor demand, establish rules and support infrastructure.
The danger is to confuse these roles. Grants should not subsidize companies forever. DFIs should not behave like charities. Governments should not pretend to be venture funds. Venture investors should not ignore unit economics. Strategic partners should not capture local companies without building local capacity.
The opportunity is to align these instruments intelligently. For African AI, the missing layer is often not money in the abstract. It is structured capital that understands the stage, risk, market, infrastructure needs and development value of the opportunity.
A pre-seed AI founder needs different capital from a data-center developer. A health-data platform needs different risk management from a consumer chatbot. An AI-enabled credit infrastructure company needs different partners from an agriculture advisory tool. A public-sector AI procurement platform needs different governance from an enterprise SaaS company.
Institutions can help define these pathways. They can create fund vehicles, guarantees, blended-finance structures, project-preparation facilities, data-governance facilities, procurement programs and regional investment platforms that crowd in private capital rather than crowd it out.
The fifth institutional role is cross-border scale
Most African markets are small when viewed individually. AI companies need customers, data diversity, regulatory clarity and revenue depth. For many founders, scale will not mean immediate continental expansion. But over time, the best companies will need pathways from one market to adjacent markets, from one sector to related sectors, or from one institutional buyer to networks of similar buyers.
Cross-border fragmentation is therefore an investability issue. Different data rules, payment systems, procurement standards, licensing requirements, digital identity systems and regulatory interpretations can make expansion slow and expensive. A company that works in one country may have to rebuild too much to work in another. That raises costs, delays growth and weakens investor confidence.
Institutions can reduce this friction. Regional bodies, central banks, data regulators, trade institutions, DFIs and standards organizations can support interoperable systems. They can promote common approaches to data governance, digital identity, payments, cybersecurity, public procurement and AI assurance. They can help make it easier for an African AI company that proves value in one market to enter another market without starting from zero.
This is where the African Continental Free Trade Area should matter. AI will not scale only through trade in goods. It will scale through digital services, data-enabled platforms, cross-border payments, logistics, compliance, language tools, enterprise software and AI-enabled professional services.
If Africa wants AI companies with regional and global relevance, institutions must help create markets large enough to support them.
The sixth institutional role is talent
Talent is not only a founder issue. It is an ecosystem issue. AI companies need engineers, product managers, data scientists, designers, clinicians, agronomists, compliance specialists, enterprise sales teams, project managers and operators. They need people who understand both technology and the sectors being transformed. They need founders who can speak to customers and investors, but they also need employees who can execute.
Universities, technical institutes, research labs, corporate training programs, AI communities and diaspora networks all matter.
The diaspora is especially important. It can bring capital, technical expertise, customer access, mentorship, global market knowledge and credibility. But diaspora engagement must be practical. It should not be reduced to speeches about giving back. It should create real channels for investment, advisory support, market access, talent circulation and institutional partnerships.
Africa’s AI talent strategy should also be honest about retention. Talented people will leave if local opportunities are weak. The best way to retain talent is not only patriotic messaging. It is to create companies, research environments, compensation pathways and ambitious missions worth staying for.
The seventh institutional role is accountability
AI can create value, but it can also create harm. Bias, privacy violations, exclusion, fraud, misinformation, surveillance abuse and unsafe deployment are real risks. If these risks are ignored, trust will erode. If regulation becomes heavy-handed, innovation will slow. Africa needs neither naïve techno-optimism nor regulatory paralysis.
It needs accountable innovation. That means clear rules, proportionate regulation, sector-specific safeguards, transparency where needed, and consequences for misuse. It also means regulators must understand the technology well enough to regulate intelligently.
Institutions must protect citizens without suffocating builders. This balance will not be easy, but it is essential. Investors will not commit serious capital into environments where rules are either absent, arbitrary or unpredictable.
The future of African AI will depend not only on what founders build, but on whether the institutional environment rewards responsible builders.
This brings the trilogy full circle
This final article argues that institutions must now help create the environment in which those companies can scale. Africa does not need more declarations disconnected from execution. It needs institutions that can build trust, organize data, finance infrastructure, open procurement, mobilize capital, enable cross-border scale, support talent and protect citizens.
The next African AI phase will not be won by founders alone, governments alone, DFIs alone, universities alone or investors alone. It will be won by those able to connect these actors into investable ecosystems.
That is why institutions matter. Not because they are more innovative than founders. They are not. They matter because they can lower the cost of trust, reduce the burden of infrastructure, create larger markets, mobilize patient capital and turn isolated promise into repeatable scale.
Africa’s AI future will be built by entrepreneurs. But it will scale only if institutions do their part.
Roger B. Jantio is an AI investor and strategic advisor focused on artificial intelligence, development finance, emerging markets, and strategic capital. He is the founder and CEO of Sterling Merchant Finance Ltd, a Washington-based merchant bank active across Africa for more than three decades, and of its affiliated investment funds.
