Bongani Ndlovu
AS a biomedical student, Ronnie Kakunguwo, an Artificial Intelligence (AI) start-up founder based in Harare, he fell in love with software engineering in his first year, starting from the very first module.
From that moment, his focus shifted toward AI and robotics.
“I worked on AI and rehabilitation engineering for my final-year research. That is where I became deeply involved in artificial intelligence and robotics. I would say my serious journey into the
AI industry began around 2023, when I started taking software development more seriously,” he tells Finance Africa.
Although Biomedical Engineering as a field of study in the southern African country focuses primarily on medical devices, he noticed a broader systemic gap.
“I noticed a significant gap in our digital health ecosystem. We do not have a well-functioning digital health infrastructure. That realisation motivated me to focus more on software engineering,” Kakunguwo said.
Kakunguwo, who is developing an AI-powered mental health system called Zyenex under his company Zyenova Technologies, says the path has been challenging.
“AI is expensive to run because it requires significant computational infrastructure. We do not have strong venture capital support like in the United States. While there are hackathons, they do not provide substantial funding for powerful AI startups,” he said.
Funding has been a major hurdle, he adds. While the United States and China are allocating billions of dollars toward data centres and AI infrastructure, Europe is mobilising sovereign capital to reduce technological dependence but the same cannot be said for Africa. To the East, countries such as India are positioning themselves as service and talent powerhouses in applied AI.
With the race for AI superiority now in full swing after Open AI introduced its large language model ChatGPT and Chinese technology companies launched competing models such as Baidu’s ER-NIE and other domestic AI systems—the question is no longer whether AI will reshape economies.
Now, Africa wants to be counted among the producers. From Kenya, Zimbabwe and South Africa, startups such as Kakunguwo’s are increasingly describing themselves as AI-first.
Some governments are drafting national AI policies. Project capital pitches often present machine learning—a subset of artificial intelligence that enables computers to learn from data, identify patterns and make decisions without explicit programming—as a multiplier across finance, health, logistics and agriculture.
For global investors, Africa represents one of the few remaining frontier markets where the foundational layers of the AI economy are still being
built. While mature markets are competing over incremental improvements in advanced models, much of Africa’s opportunity lies in constructing the underlying digital infrastructure—data centres, fibre backbones, GPU clusters and power systems—that will support the continent’s next generation of AI companies.
This creates space not only for venture capital backing startups, but also for infrastructure funds, sovereign investors and technology partners seeking long-term exposure to emerging digital economies.
Against such a background, Africa does not want to be left behind.
However, for a continent grappling with unreliable electricity supplies, limited data centre density and uneven broadband penetration, the AI ambition may remain just that—an ambition.
Unlike fintech, Africa’s most prominent technology success story, AI is not reliant on minimal infrastructure. Fintech companies like Ecocash, TX Money Transfer, Mukuru, Innbucks and M-Pesa have scaled by utilising existing mobile networks and relatively modest cloud infrastructure.
Their revenue models, which are based on transactions, allowed for swift monetisation, clear cash-flow visibility and relatively low initial capital investments. Owing to these favourable factors, startups were able to scale rapidly, expand across borders and connect with millions of users without needing to invest heavily in physical infrastructure upfront.
The contrast between fintech’s rapid growth and AI’s infrastructure demands also reveals a potential investment pathway. Just as early telecommunications expansion unlocked Africa’s mobile money revolution, the build-out of regional computing capacity, affordable broadband and reliable power could catalyse a new wave of AI companies.
Investors who finance these enabling layers stand to benefit from both infrastructure returns and equity exposure to the startups that emerge on top
of them. In the view of Deep Analytics Chief AI Scientist Panashe Chiurunge, Africa can build AI companies under current infrastructure conditions—but
not at scale and not across the full technology stack.
For him, application-layer AI startups are viable today. Firms focused on fintech scoring, fraud detection, agricultural advisory systems, African language models or health diagnostics can operate using global cloud platforms, he says.
He points to InstaDeep, which built advanced AI systems in Tunisia before being acquired by BioNTech, while DataProphet continues to deploy machine learning solutions in global manufacturing environments. In his view, these companies demonstrate that Africa can compete at the application and research layers despite infrastructure constraints.
But Chiurunge warns that Africa risks repeating historical patterns of extractive participation in technological revolutions unless it addresses foundational deficits.
“In previous industrial revolutions, Africa exported raw inputs. During the first industrial revolution it supplied cotton. In the second, coal and minerals. In the digital and AI era, it supplies cobalt, lithium and rare earths. Without reliable
power, high-density data centres, graphics processing units (GPU) clusters and broadband backbones, Africa risks repeating that extractive role,” he told Finance Africa.
“If you build the ‘cart’ of AI applications before the ‘code’ of infrastructure and skills, you lock yourself into dependency. Competitive AI ecosystems require stable electricity, local data centre capacity, affordable broadband and advanced STEM and AI research pipelines. Application startups can survive. Foundational AI leadership requires infrastructure reform,” he said.
For foreign investors, these gaps are not only constraints but entry points.
Large-scale investments in renewable-powered data centres, regional cloud infrastructure, AI training facilities and high-performance computing clusters could anchor new digital ecosystems across the continent.
With Africa’s population projected to exceed 2.5 billion by 2050 and mobile connectivity continuing to expand, demand for AI-driven services in finance, healthcare, agriculture and logistics is expected to grow significantly.
In the context of AI, Africa currently provides crucial resources, including cobalt, lithium and rare earth elements that underpin global GPU supply
chains.
However, without reliable energy supplies, robust data centres, GPU clusters, broadband infrastructure and advanced STEM research capabilities, Africa may continue exporting raw inputs while importing AI-driven solutions and intelligence.
Chiurunge argues that while local AI application startups may emerge, true AI leadership and infrastructure development depend on overcoming these foundational challenges and establishing sovereign data governance frameworks.
In his assessment, African AI startups rely heavily on offshore cloud infrastructure and are exposed to high costs due to dollar-based pricing, reduced control over sensitive data and scalability limits imposed by foreign providers.
While offshore cloud services are useful in the short term, overreliance risks making African firms “renters” in the AI economy, he says.
Instead, he advocates a hybrid strategy—leveraging global cloud services for speed while building regional GPU clusters and sovereign infrastructure for long-term leadership.
Grace Murugi, Chief AI Strategist at This is Digital, insists Africa must advance infrastructure and innovation together rather than sequentially.
For her, Cassava Technologies’ partnership with NVIDIA to construct AI factories in several African countries—including South Africa, Nigeria, Kenya, Egypt and Morocco—to boost local computing power and reduce reliance on foreign GPU infrastructure is a case in point.
Murugi notes that while foreign cloud services allow for rapid AI deployment, they increase costs, expose startups to currency risks, cause high latency and complicate compliance with data sovereignty laws when sensitive data is stored offshore. Looking ahead, Murugi sees three main opportunities for Africa to build lasting value over the coming decade.
These include owning and developing rich local datasets—especially in African languages and key sectors—creating applied AI solutions for local markets such as agriculture, education, logistics and mobile finance, and investing in digital infrastructure, including regional data centres and energy-efficient computing.
Focusing on these areas will help Africa secure a competitive and sustainable position in the global AI landscape, she says.
Scotel Private Limited Managing Director Steven Mashingaidze says there is an urgent need to develop infrastructure alongside innovation.
He points to progress in data centre projects powered by renewable energy but notes that limited capacity, high data costs and uneven broadband coverage remain significant obstacles.
He highlights the importance of digitising and integrating national datasets as a foundation for AI, suggesting that universities could host distributed data centres to boost local computing capacity. Mashingaidze also sees Africa’s mineral wealth as a potential asset for future semiconductor and GPU manufacturing, though global dependencies persist. (Source: Finance Africa Quarterly, a Bard Global Finance Institution publication)
Online Reporter