How Africa can re-imagine its artificial intelligence future

Much has been said about artificial intelligence (AI) in Africa. But the dominant questions are: Will Africa be left behind as countries at the frontier of AI development consolidate economic and productivity advantages? What foundations must be built for AI to be adopted and scaled in the first place? And how can AI enable African countries to leapfrog traditional development pathways?

What has remained largely unspoken is that partnerships must be re-imagined so that AI diffusion delivers real, people-centred impact and, in turn, enables Africa to systematically meet the prerequisites for AI adoption at scale. In the age of AI, the design of partnerships is not a secondary consideration. It is the central determinant of whether innovation translates into inclusive, sustainable development.

As the world convenes in New Delhi this February for the AI Impact Summit, the first major global AI gathering hosted in the Global South, the Nairobi AI Forum emerges as a critical precursor. It shifts the global conversation from aspiration to execution and from broad principles to concrete systems, focusing on what must be built, how it must be built, and with whom.

The Nairobi AI Forum builds on the momentum of the AI Hub for Sustainable Development, which was launched under Italy’s G7 Presidency and the Mattei Plan. The AI Hub is powered by the Ministry of Enterprises and Made in Italy and implemented by UNDP, in partnership with the G7 and Kenya. Unlike traditional conferences, the Forum is an intensive, action-oriented working session that forges market-based, non-Official Development Assistance partnerships among African innovators, governments, and private-sector leaders from the G7 and the European Union.

This is why the forum is organised into three tightly linked working blocks: Unpacking AI infrastructure in the African context, building partnerships across borders, and unlocking financing. Together, these blocks form a coherent pathway to translate technical ambition into investable, scalable systems.

At the centre of this approach is a simple yet often overlooked reality: AI adoption requires horizontal enablers shared capabilities that span sectors and use cases. These include reliable power, high-speed connectivity, affordable compute, sovereign data systems, robust safeguards, and deep talent pipelines. Without these foundational layers, AI diffusion remains fragmented, costly, and confined to isolated pilots.

At stake is not merely technological capability but sovereignty in the age of intelligent systems. For African countries, building sovereign AI capacity does not imply technological isolationism. Rather, it means ensuring strategic independence in an era of deep interdependence, owning critical layers of the AI stack while remaining interoperable with global systems, standards, and markets. Without this balance, nations risk outsourcing not only infrastructure but also decision-making capability itself. Sovereign AI, in this sense, becomes a foundation for democratic accountability, economic resilience, and long-term autonomy in development.

Yet prevailing global infrastructure models have often failed to reflect African realities. Hyperscale data centres designed for other contexts often require hundreds of megawatts of continuous power, far exceeding available grid capacity in many countries. The result has been costly infrastructure with low utilisation, limited spillovers into local innovation ecosystems, and weak alignment with national development priorities.

The Nairobi AI Forum centres on a different approach: right-sized, integrated AI ecosystems aligned with local energy availability, workloads, and talent, and grounded in Africa’s vast renewable energy potential and strategic critical-mineral endowments. Rather than pursuing capital-intensive hyperscale deployments, the focus is on modular, distributed systems that scale with actual demand and deliver immediate economic and social returns.

As Kate Kallot, a leading voice in the AI Hub’s Private Sector Advisory Group and a recent TIME 100 honoree, has argued: “Africa must reject the ‘billion-dollar hyperscale trap’ and instead build modular, sovereign infrastructure containerised Graphics Processing Unit (GPU) clusters matched to real demand and power availability, embedded within a full stack that includes connectivity, sovereign data pipelines, safeguards, and affordable compute.”

This requires building across the full AI stack from data generation, stewardship, and governance to compute infrastructure, skills development, model adaptation, and real-world use cases. Investment must therefore flow not only into hardware but also into sovereign data pipelines, public-interest model development, regulatory and safety institutions, and large-scale workforce skilling. Only by advancing these layers together can African countries ensure that AI adoption translates into sustainable economic transformation rather than technological dependency.

When infrastructure is designed this way, partnerships begin to operate differently. Innovation becomes grounded in deployment, and investment follows demonstrated demand rather than speculative scale. This dynamic is already evident in the AI Hub’s flagship programmes, including the AI Infrastructure Builder Program and the Compute Accelerator Program, where real-world deployments guide both technology design and capital allocation.

Innovators such as DeepLeaf, emerging from partnerships between Morocco and Italy and now expanding into Kenya, and Crane AI, scaling from Africa into the United Arab Emirates, demonstrate how context-driven collaboration accelerates adoption. Practical deployment reveals precise technical and market needs, creating powerful feedback loops among entrepreneurs, investors, infrastructure providers, and governments. From food systems and climate resilience to healthcare delivery and public financial management, these systems anchor AI not in abstraction but in everyday development outcomes.

Mr Vincenzo is the ambassador designate of Italy to Kenya and Seychelles and Permanent Representative to UNEP and UN-Habitat. Mr Thigo is Kenya's Special Envoy on Technology. Ms Keyzom is the Head of Digital Programming, UNDP Chief Digital Office.