Four leaders on what Africa already knows about AI, and what it does with that next

Ask the global AI industry where Africa sits on its map, and you will usually get one of two answers: a market to be reached, or a gap to be closed. Ask four people actually building with AI on the continent, and you get something else entirely.

Meet the panel:

• Caitlin Burton , CEO, Zipline Africa
• Charles Murito , Regional Director, Government Affairs and Public Policy, Sub-Saharan Africa, Google
• Lavina Ramkissoon (Amb. Dr. ) , Ambassador and Co-Chair, AU-ASRIC and AU-Sixth Region, African Union
• Joost Pielage, Chief Technology Officer, Quro Medical

We gave each of them ten minutes and the same five questions: what people get wrong about AI in Africa, where it can have the greatest impact, what it will take for the continent to move from adoption to leadership, what they're watching most closely, and what role events like Africa Tech Festival play in getting there. Here is what they told us.

The catching-up myth

Every one of them starts by rejecting the premise that Africa is behind.

Burton's case is the most concrete, because it has been running the longest.

I think the thing people are missing is that in some areas, Africa is not behind on AI at all. It is already showing the rest of the world what happens when AI moves out of a computer and becomes infrastructure.

Murito's answer starts from the arithmetic of service delivery. "Many view Africa merely as an AI consumer market or worry about an expanding digital divide," he says. "What others miss is that African innovators are deploying AI to solve the continent's acute ratio problem: the severe shortage of doctors to patients, teachers to students, and agricultural extension workers to smallholders. In Africa, AI is not a luxury of convenience; it is an essential multiplier of human capacity."

He argues the constraints have shaped the builders rather than held them back. "Faced with infrastructure and compute constraints, local developers are pioneering frugal, offline-friendly innovations out of necessity. Paired with having the world's youngest and fastest-growing population, Africa is actively cultivating the homegrown talent that will drive the future global AI economy."

Ramkissoon takes issue with the tense.

My optimism about AI in Africa comes from the same place as my optimism about Africa itself: our people. We should approach it with curiosity, ambition and responsibility.

Pielage moves the question inside the engineering team, and lands somewhere the product roadmap rarely reaches. "Most companies look at their product and ask themselves, where do we put AI in our product? The better question is, where do we put AI in how we build the product?" Not every product needs tons of AI features, he says, most products do perfectly well with normal logic, but every engineering team benefits from incorporating AI into their SDLC. "That's where the leverage is much bigger."

The reason matters for anyone hiring on the continent. "Our constraint in Africa isn't ideas; it's senior engineering capacity. If AI lets a team of four engineers create products like a team of twenty, that changes what a company can compete for. Product AI is visible. SDLC AI is compounding."

Where intelligence has to act

Ask four leaders where AI matters most, and you might expect four different answers. What comes back instead is four versions of the same gap: a system that knows exactly what it needs and cannot do anything about it.

"I think the biggest opportunity is where the constraint is not a lack of information, but the inability of a system to respond," Burton says.

Healthcare makes that very tangible. You can have excellent demand forecasting, patient data and clinical decision support, but none of it helps the patient if the medicine, vaccine or blood they need is sitting somewhere else. That is where combining AI with robotics becomes powerful. You can understand demand, manage inventory more intelligently and then physically respond to that demand in real time.

Murito applies the same logic to classrooms, and opens with a number that sets the scale. "AI can incredibly help the learning poverty challenge. 89% of 10 year olds in Africa currently struggle to read at their proficiency level. Using AI, we can help bridge the gap by delivering more personalised learning tools like Google's Read Along to help young learners engage at their own pace, even in resource-constrained environments."

From there it widens. "Beyond education, AI can revolutionize healthcare diagnostics, optimize food security for smallholder farmers, and empower the small and medium enterprises that power over 80% of our economies. By personalizing foundational learning and closing critical service gaps, AI turns historical resource deficits into engines of inclusive growth."

For Ramkissoon, the bottleneck is administrative rather than technical. "That point is reached when AI stops being a geo-political or technology story and becomes an impact story. Where people are solving real problems, often in survival mode. We have been survival-coding for the last few decades."

Fix the paperwork and the economics follow: "One of its biggest economic opportunities is breaking bureaucratic bottlenecks: lowering costs, lowering processing time, lowering barriers of entry while increasing the speed of coordinated real-time insights to increasing continental GDP. That means lower costs, greater productivity, more trade and more opportunity. Simply creating value, moving faster, with purpose on our own terms."

Pielage returns to a word Africa has heard for twenty years and explains why it finally means something different.

The usual answer is leapfrogging. For the past twenty years, the sequence has been: digitise first, then automate. AI breaks that sequence. Models handle unstructured inputs extremely well, voice notes, handwriting, WhatsApp messages, and photos, for example. As a result, you no longer need to rebuild every process around perfectly clean, structured data before you can automate it. Removing that most expensive step creates an opportunity to leapfrog into the future.

Who owns the value

This is where the panel gets deliberately uncomfortable. Each of them refuses the definition of leadership currently on offer.

I would challenge the idea that AI leadership can only be measured by who builds the biggest model.

That's Burton's full answer, and it sets the terms for the other three.

Pielage draws the same line from inside a company solving a problem richer health systems have never faced. "Leadership isn't training foundation models, that race is expensive and largely settled. It's being the best in the world at applying them to problems we understand better than anyone else in our own context and region." For him that means delivering the best hospital-level care in a home setting, in a country with a fraction of the specialists per capita. "This is a harder problem than most US health systems have ever had to solve."

Solve it once and it travels: "Solve it in South Africa and it exports across most African countries. Leadership is when the rest of the world starts importing African implementations, not just African data."

Ramkissoon asks the question the sector tends to leave until last.

We should also be willing to ask a slightly uncomfortable question: who owns the value created by AI? Thus we need to stop being polite guests at the global geo-technological fronts. Trust needs to be regained amongst member states, technology must be placed at the centre of the AU agenda, dependency needs to be removed and lastly a bold visionary roadmap for all.

Murito sets out what the policy layer has to deliver for any of that to hold. "To move from adoption to leadership, Africa must transition from passive consumption to active, genuine collaboration across the entire innovation ecosystem. First, governments need forward-leaning, pro-innovation policies. This means adopting cloud-first frameworks, facilitating cross-border regulatory sandboxes under the AfCFTA, and opening public sector datasets for collaborative research and development."

Data comes next, and he frames it as national infrastructure rather than raw material. "We must treat African language and public data as shared, sovereign infrastructure through deep partnerships between academia, industry, and local builders. Most crucially, leadership requires sustained investment in human capital: moving beyond basic digital literacy to nurture skilled AI engineers and innovators."

On the watchlist

What they are each tracking says as much about their priorities as anything else.

Murito is watching models small enough to run anywhere, and speech technology built for how people actually speak. "I am watching the rapid evolution of lightweight, offline-capable open models and distributed edge computing that democratize access while preserving data sovereignty. Alongside this, I am closely tracking advances in localized voice and language technologies, exemplified by our Project Waxal initiative." By partnering with African academic institutions to build open speech datasets across more than 20 local languages, he says, Waxal proves that capturing authentic cadence and indigenous dialects is foundational to digital inclusion. "The most transformative AI for Africa is not necessarily the largest model, but the most relevant and accessible one."

Pielage is watching the bill.

What I am watching most closely is the pace of model development and the broader direction in which the technology is evolving. The capabilities of leading models continue to improve rapidly, but the more important question is whether those improvements translate into meaningful real-world value. I am also paying close attention to the economics of AI, particularly token pricing and the overall cost of running AI-powered solutions at scale. Ultimately, the key issue is sustainability: can these technologies continue to advance while becoming more affordable, and can organisations deploy them in a way that delivers measurable value while remaining cost-effective over the long term?

Ramkissoon's list runs wider and further ahead. "I am particularly interested in the movement of balance between man and machine, especially with ASI close to reality. Geo-politics most definitely. Science arena of AI (physics, biology, longevity), psychology of youth, quantum, and defence technologies are up the list. Other areas include Edge AI, security (cyber, privacy, online safety)." What connects them, she says, is the point at which "true abundance and democratisation happens."

What the room is for

The panel's final question is about events like this one, and their answers amount to a standard to be held to.

Platforms like Africa Tech Festival are valuable because building Africa's AI future will require far more than technology companies working in isolation.

Burton, again, sets the tone for what follows.

Ramkissoon wants the room to argue. "The most powerful thing about the Africa Tech Festival is bringing the right people into the same room, and then making sure they don't leave with only great selfies and business cards. It is the crucible where policy meets capital. We need the regulators, the hackers, and the investors in the same room, arguing, collaborating, and signing deals." The output she is looking for is a change in register: "The festival's role is to turn the global Africa hype into African trusted ecosystems. Africa has extraordinary innovators, researchers, entrepreneurs, investors and policymakers. What we need is more connection, collaboration and conversion of ideas into action."

Pielage makes the practical case. Shared experience stops teams paying twice for the same mistake. "It's important to share the latest developments and create opportunities for people from across the continent to come together, exchange ideas, and learn from one another's experiences. By doing so, we can avoid repeating the same mistakes, accelerate progress, and build on what has already been proven to work in different markets and industries." And the imported playbook rarely fits: "As with most technologies, solutions developed primarily for the US market are not always well suited to the realities and needs of African businesses and communities."

Murito sets the longest clock on it.

Africa Tech Festival provides an unparalleled platform to bridge the divide between policymakers, entrepreneurs, enterprise leaders, and civil society. AI cannot advance in isolation; it requires deep alignment across national policy, digital infrastructure, education and skills development and capital investment. Crucially, this collaboration cannot begin and end during a single conference week. We need to ensure that the advocacy and partnerships continue well beyond the week of the festival itself. The platform must serve as a year-round catalyst, translating stage dialogues into concrete policy harmonization, cross-border collaboration, and measurable action plans. The true test of convenings like Africa Tech Festival is not what happens on stage, but the sustained advocacy and action that continues long after the delegates depart.

Join the conversation at Africa Tech Festival this November and be part of shaping Africa’s digital resilience. Together, we can build a secure and innovative future for the continent. Register now to secure your place!

Africa Tech Festival returns to the CTICC in Cape Town from 17 to 19 November 2026. Caitlin Burton, Charles Murito, Dr Lavina Ramkissoon and Joost Pielage all take part in this year's AI programme.