In a recent episode of Futurise, Shikoh Gitau, CEO of QHALA, reframed the global AI conversation through an African lens. She gave a point of view in which relevance, access and live impact matter far more than model speed or benchmark scores.
AI as a Social Infrastructure: Why Impact Matters More Than Models in Africa
Shikoh describes AI as a consequential technology, drawing a powerful parallel with the mobile revolution that transformed the African continent between 2007 and 2010. Mobile phones failed because of technical elegance; they succeeded because they removed barriers such as cost, access and dependency. AI, she argues, sits at a similar inflection point today.
For Africa’s 1.5 billion people, the value of AI is not theoretical. It is deeply practical. Can it help educate a child? Provide medical guidance? Improve access to justice? Or support livelihoods? These are the questions that determine adoption. In this context, AI becomes less a product and more a social infrastructure. It can act as a means of democratising knowledge previously locked behind paywalls, institutions or geography.
Crucially, Shikoh highlights that the last mile matters the most. While frontier models dominate headlines in the Global North. Most of Africa’s population remains disconnected from solutions that translate AI capability into daily life. Making AI real means embedding it in familiar channels, local languages and culturally understood forms of interaction — not abstract interfaces designed elsewhere.
Building an AI Ecosystem, Not Just AI Tools
Shikoh outlines five foundational pillars required for a sustainable African AI ecosystem: governance, data, compute, use cases and talent. Each exposes structural gaps that cannot be solved by technology alone.
Africa contributes less than 2% of global AI training data, despite its vast linguistic and cultural diversity. Thousands of African languages remain undigitised. Interestingly, even digitised ones often reflect formal or colonial versions rather than how people speak. This creates exclusion at the model level. Solving this requires digitisation, aggregation, labelling and contextualisation. This makes it a long-term investment, not a quick win.
Compute presents another constraint. With limited high-performance infrastructure on the continent, African organisations struggle to train models locally or experiment at scale. This reinforces dependency and slows innovation. Meanwhile, talent is unevenly distributed: Africa supplies a significant share of the world’s data annotators yet contributes less than 1% of AI research output. Closing this gap requires deliberate investment in skills across the full spectrum. Right from the data collection to advanced research.
What distinguishes QHALA’s work is its focus on use cases with immediate social relevance. Whether supporting women’s health through culturally grounded AI companions or improving access to legal guidance, these applications demonstrate a dual return: measurable social impact alongside economic viability.
The message is clear: In Africa, people will adopt AI not because it is new, but because it is trustworthy, useful, and makes a real difference. Success will come from building systems that meet people’s needs where they are, rather than expecting them to adapt to the technology.
