Stephen Lamprecht, founder of South African management consultancy VS Nova and a research partner on the survey.
While ICT accounted for 62.7% of all southern African venture capital deals in 2025, AI and machine learning represented just 1.6% as a standalone category.
This is according to the Southern African Venture Capital and Private Equity Association’s (SAVCA) 2026 Venture Capital Industry Survey.
Stephen Lamprecht, founder of South African management consultancy VS Nova and a research partner on the survey, said the figure understates AI’s influence because the technology is increasingly embedded across sectors rather than confined to a standalone investment category.
“AI is everywhere,” said Lamprecht. “It’s part of every sector and every industry, whether that’s healthcare, financial services, agriculture or ICT itself. We can no longer treat AI as a separate category because increasingly it’s being absorbed into the application areas where it creates value. The real story is not about standalone AI companies, but about how AI is enabling innovation across the economy.”
At the SAVCA VC Conference in Stellenbosch in September, Jason Goldberg, co-founder of South African enterprise and supplier development specialist Edge Growth, said many organisations underestimate the scale of change AI is likely to bring.
“Many people think AI is a fun tool to write e-mails faster, create slide decks or improve productivity. But it is much more than that. AI is a general-purpose technology, like the steam engine, electricity or the internet. It changes what becomes possible.”
Goldberg said the biggest risk for business leaders and investors is assessing an exponential technology through a linear lens.
“When the curve is exponential and your paradigm is linear, that's when you lose,” he added. “The companies that thrive are the ones that recognise how quickly the economics of an industry can change when a new technology fundamentally alters what is possible.”
He argued that AI is making expertise and problem-solving more accessible, much as electricity transformed access to energy. “The difficulty of solving complex problems is what AI is fundamentally disrupting.”
Goldberg urged investors to assess whether AI will increase or reduce the value of a company’s offering over time.
“As investors, the key question is simple: what becomes more valuable as AI gets better, and what becomes less valuable as AI gets better?”
He pointed to declining valuations in parts of the software sector as an example. AI is making software development faster, cheaper and more accessible, lowering barriers that previously required specialised skills, large teams and significant capital.
“AI is making some things that used to be hard, expensive and valuable become cheap and easy. That's a primary lens we need to apply as investors.”
New scale advantage
Goldberg identified emerging opportunities in AI-native services businesses, where expertise can be embedded in agentic systems to deliver services at lower cost and with less complexity.
“For years, venture investors largely avoided professional services businesses because they were difficult to scale,” he said. “Today, expertise itself can increasingly be embedded into systems that deliver high-value outcomes faster, more consistently and at significantly lower cost. That fundamentally changes the economics of those businesses.”
He described this as a “new scale advantage”, enabling organisations to access capabilities previously constrained by specialised talent, management capacity or institutional knowledge.
For SA’s venture capital ecosystem, AI’s limited visibility as a standalone category reflects its growing integration across industries.
Johan Steyn, AI expert and founder of AIforBusiness.net, said the 1.6% figure is accurate but is the least interesting number in the survey.
“AI is becoming a horizontal layer that runs through almost every deal – fintech, health tech, agritech, logistics – rather than a category of its own, so counting 'AI deals' separately today is a little like counting 'internet deals' in 2005: within a few years, the label becomes meaningless because it is in everything. South Africa's real opportunity is not in funding frontier models we cannot afford to build, but in applied AI that solves hard local problems using data and context that do not port from Silicon Valley.
“Whether we actually capture that value depends on the fundamentals – clearer policy, a deeper local capital pool, and keeping our best companies and talent domiciled here rather than re-domiciling offshore the moment they raise,” he added.
Lamprecht said AI is making businesses scalable that previously struggled to deliver the returns venture capital investors require.
"This is happening as we speak and basically undoing some of the 'best practices' of VC investment," he said. "But it takes time, because you have to ask: is it real, will these new trends prevail? It is too early to say."
What has not changed, he stressed, is the VC investment model itself. "Investors still need very high returns to offset losses across a high-risk portfolio, typically within three to seven years to meet their investors' requirements. AI may create new opportunities, but those businesses must still deliver returns within that model."
He also cautioned that AI is not limited to local start-ups. "Keep in mind, AI is not limited to our start-ups; their competitors also use AI."
Lamprecht said future investment opportunities will increasingly centre on businesses using AI to improve efficiency, solve problems differently and create value across the economy.
“Its influence has grown so pervasive that it is increasingly disappearing as a category altogether, becoming embedded in the technologies, products and business models that will shape the next generation of venture-backed growth companies.”
Investment opportunities remain uncertain
Lamprecht said it remains unclear where the next generation of AI-enabled investment opportunities will emerge.
“The start-up, tech and VC environment has always been dynamic, and we are possibly at the beginning of a global rewriting of the script on investment opportunities," he said. "But we have not yet gone through a full cycle of AI-driven VC investments through to exits. The upside will probably materialise given that AI is a productivity enabler, creating opportunities for growth and better returns."
Lamprecht said he would like to see how start-ups are using AI's scalability to address opportunities that were previously not commercially viable, like using AI in public healthcare, or solving public finance oversight through a blockchain mechanism that pays for itself.
Beyond the current "obvious" digitisation cycle, Lamprecht believes more opportunities will come from critical service delivery needs that were previously off-limits because no commercial return model could be found.
"An example of this is how the likes of Netcare and Dischem are clearly positioning themselves in the public health space as a solution for NHI – this would not have been thought possible even five years ago," he said. "Public health is a huge spending area of government and could open up enormous investment opportunities, but not with government as the gatekeeper – rather with government’s proxies, like those private sector players. I think this is very exciting.”
He cautioned that technology alone does not determine investment success.
“Venture capital is about investing in people and businesses, not technology," he said. "This year’s SAVCA survey again reiterated the value-add of investors beyond capital – advice, governance, networking and even operationally getting hands-on. As AI accelerates growth and increases uncertainty, experienced investors will need to help founders turn technological potential into sustainable businesses, regardless of sector, public or private. The human and commercial support will remain, or even become more critical.”