FEATURE
Werner Joubert, Commercial SYS Business Director for South Africa and SADC at ASUS
Patrick Dupoux, MD and Senior Partner at BCG
Troye Technical Director Kurt Goodall
Unlike mature economies burdened by decades of legacy technology, African businesses often have the freedom to build AI-first environments from the outset.
Werner Joubert, Commercial SYS Business Director for South Africa and SADC at ASUS, agrees.
“ The continent is home to a young workforce,” he said.“ Africa has significantly less legacy infrastructure that needs to be retrofitted to accommodate new technologies.”
That allows startups and enterprises alike to adopt AI-native approaches much faster than organisations still modernising decades-old IT estates.
Infrastructure becomes the next competitive advantage
While enthusiasm around AI continues to grow, industry pundits identify one issue that will ultimately determine Africa’ s long-term success: infrastructure.
Artificial Intelligence requires enormous computing resources. Without sufficient cloud capacity, GPU infrastructure, connectivity and local data centres, organisations risk becoming permanent consumers of technologies developed elsewhere.
Patrick Dupoux, MD and Senior Partner at BCG, points to one striking statistic: Although Africa represents approximately 18 % of the world’ s population, it accounts for less than 1 % of global data centre capacity.
That imbalance limits not only where AI models can be trained and deployed, but also where economic value is ultimately retained. Rather than expecting every nation to build its own AI ecosystem independently, BCG advocates regional collaboration.
Shared investments in cloud infrastructure, digital identity platforms, high-performance computing and core digital services would allow countries to pool resources while maintaining strategic control over data and governance.
Public-private partnerships, supported by governments and development institutions, will therefore become essential to Africa’ s digital future.
Troye Technical Director Kurt Goodall agrees infrastructure matters but challenges the conventional definition of AI sovereignty. Instead of attempting to compete with hyperscale nations building frontier AI models, he argues Africa should prioritise controlling the layers above them.
That means embracing open-source and open-weight AI models, hosting them locally, securing them properly and building differentiated applications on top of proven global technologies.
For Goodall, sovereignty is less about constructing every component and more about maintaining control over how AI is deployed and governed.
Owning the data, not just the technology
Infrastructure alone, however, is only part of the equation. If Africa is to build truly sovereign AI capabilities, many experts argue that it must also own the data, skills and intellectual property that underpin artificial intelligence.
The debate is not necessarily about creating an African equivalent of every large language model. Instead, it centres on ensuring AI systems genuinely understand African realities.
For Maher, the continent has reached a point where simply adopting global technologies is no longer enough.
“ Africa must increasingly build, govern and own the infrastructure, data, capabilities and innovation ecosystems underpinning AI,” he says.
Today, Africa’ s digital economy accounts for just 5 % of GDP, compared with a global average of 15 %. While that figure is projected to grow, BCG argues that the continent’ s next phase of development must focus on producing technology rather than simply consuming it. Ownership of AI infrastructure and ecosystems will determine where future economic value is created.
Ergin shares a similar perspective but believes the discussion should be framed differently.
“ The objective isn’ t to build an African version of every large language model that already exists. The objective is to ensure AI genuinely understands Africa.”
That requires investment in African languages, regional datasets, regulatory frameworks and local expertise so that AI systems reflect the continent’ s cultural and economic diversity.
“ I don’ t see this as a choice between global AI and African AI,” he said.“ I see the future as
26
INTELLIGENT CIO AFRICA www. intelligentcio. com