A recent broadcast brought together three people working on the ground: Salma Kabbaj, secretary general of Afrique Innove and founder of Impact Lab; Charles Burkinda, co-founder and CEO of Digitech Africa; and Naufal Chama, founder and president of Startup Maroc. Their answers paint a more nuanced picture than the usual talk of an AI “revolution.”
A real acceleration and greater autonomy
On one point, all three speakers agree: AI tools save time and make entrepreneurs more agile. Charles Burkinda describes a democratization of access to analysis. Comparing business models, preparing a negotiation or producing a first draft of a document are tasks that once fell to specialist firms and that, in his words, can now be done in-house in a matter of hours.
His company operates in several countries, facing different regulators. AI allows him to analyze and compare regulatory texts without systematically relying on outside partners. He also notes a shift in the skills being sought: fewer developers in the traditional sense, and more “architects” capable of supervising AI agents that write the code.
For her part, Salma Kabbaj mentions first-level legal support that is available earlier in the fundraising process, and far easier access to market data than before. She warns, however, that these gains only hold if the entrepreneur understands what they are doing.
Applications that all look alike
This is perhaps the most striking observation of the broadcast, and it comes from the incubator side. Naufal Chama explains that the quality of applications has improved markedly: polished business models, three-year projections, clean files. The problem is that this façade of quality no longer makes it possible to tell a good project from a bad one since, as he puts it, all the files are written by AI and look the same.
Salma Kabbaj makes the same observation during calls for applications to her programs. She describes texts with flawless sentences, whereas levels used to vary widely. Charles Burkinda sees a broader risk: the standardization of offerings, which makes differentiation harder.
Incubators are therefore adjusting their criteria. They no longer judge the file; they question the fieldwork. Naufal Chama sums up the new question put to candidates: how many customers have you met, and what did they tell you? An AI can write an application, he points out, but it cannot interview customers on the founder’s behalf.
The role of incubators is changing
If information is now available to everyone, what is an incubator for? The speakers answer with a change in the nature of support. Naufal Chama explains that these structures are no longer “content providers” but providers of connections to the field. AI also serves incubators themselves, helping them analyze the applications they receive in depth, spot weaknesses and propose implementation plans.
Charles Burkinda follows the same reasoning from the entrepreneur’s point of view. If he were starting a company today, he would still join an incubator, but not to get a business plan or training in fundraising. He would look for access: the meeting with a prospect he could not get on his own, the first customer, the credibility needed to pass a credit committee.
Adoption nevertheless remains uneven. According to Salma Kabbaj, entrepreneurs embraced these tools very quickly, while support organizations are moving more hesitantly, often on the initiative of individuals. Very few have developed AI agents tailored to their line of work. This is one of the themes of the upcoming Afrique Innove summit.
Does Africa need its own AI?
The question of technological sovereignty divides the speakers less than it adds nuance. Naufal Chama considers the idea of a “made in Africa” AI to be “distant” given the sums committed by the major global players, which he estimates at thousands of billions of dollars. He highlights another point: an entrepreneur in Dakar, Abidjan or Casablanca has access to the same level of technology as a counterpart in Silicon Valley, Shanghai or Paris. A rare advantage, in his view, since few industries offer such a level playing field at the outset.
Charles Burkinda points to the investment gaps between the United States, China and Europe, and cites the example of high-performing Chinese models developed at lower cost. For him, the African challenge is one of highly specialized skills, capable of designing optimized, economical solutions. He also suggests pooling efforts at regional level, and envisages the use of open-source AI infrastructure that companies could train on their own data to address local problems.
Salma Kabbaj shifts the debate: the opportunity lies less in the technology than in its uses. She observes that African entrepreneurs are often driven by the need to solve a critical problem they see around them. She cites agritech solutions that detect crop diseases early and optimize irrigation, as well as a startup she supported that automates the analysis of ultrasound scans of pregnant women to spot certain conditions sooner.
Funding remains the sinew of war
None of this removes the question of resources. Naufal Chama points out that Moroccan incubators do not earn revenue from startups, either through program fees or equity stakes. They depend on public programs, donors and international cooperation bodies, which creates sharp disparities. His organization can, he says, award grants of up to $20,000 per startup and manage funding from a Dutch agency amounting to $35,000 to $50,000 in total, but he stresses that this is not the norm. Many applications are turned down, not because they are weak, but because there are not enough resources to fund them properly.
He calls for long-term partnerships, since one-year programs prevent incubators from structuring themselves. Salma Kabbaj adds that startup success rates are low everywhere in the world. A much larger pool of projects must therefore be fed, in her view, for a significant number of continental champions to emerge.
Charles Burkinda opens up another avenue: public procurement. He recalls that, according to him, Africa captures less than 1% of global funds, and believes that a state that wants an AI sector must first buy local. Major public digital contracts are still rarely won by companies from the sub-region, which he sees as a political choice rather than a technical inevitability.
Countries are moving at different speeds
The reports aired during the broadcast give a glimpse of this diversity. A few points of reference, as relayed by the correspondents:
- in Senegal, a ten-year plan called the New Technological Deal structures the digital strategy, with challenges identified around funding, training, computing capacity and data protection;
- in Gabon, the document flows of the timber and oil sectors are presented as fertile ground for automation, and a framework for personal data protection exists;
- in Nigeria, incubators and innovation hubs, including those of Lagos State, are building on initiatives such as the one launched in April at the University of Lagos;
- in Côte d’Ivoire, banks, telecoms and retail are adopting AI, but a shortage of qualified staff mainly holds back SMEs.
As AI makes document production commonplace, value is shifting toward what it cannot do: meeting customers, convincing a first buyer, understanding one’s own profitability. African incubators, for their part, are being asked to reinvent themselves, with funding that remains fragile. The Afrique Innove summit, scheduled for November 6 and 7 in Lomé, Togo, is set to address the quality of support, the funding of support organizations, AI tools and cooperation with public authorities. It remains to be seen whether its recommendations will translate into lasting commitments.
FAQ
Is AI replacing incubators in Africa?
No, according to the speakers. It changes their role: less content and training, more connections with customers, partners and funders.
Why do application files look so alike?
Because they are increasingly written with AI tools. Incubators are therefore adjusting their criteria and placing more emphasis on validation in the field.
Can Africa develop its own AI models?
The speakers consider it difficult in the short term, given the sums involved. They favor local solutions built on existing technologies, and mention regional or open-source approaches.
Where will the ecosystem’s next major gathering take place?
The Afrique Innove summit will take place on November 6 and 7 in Lomé, Togo, according to Salma Kabbaj.
