.png)
De Nairobi à Lagos, en passant par Kigali et Accra, une nouvelle génération de fondateurs développe des innovations et des infrastructures en IA qui sont au cœur du quotidien de millions de personnes : accès aux soins, transferts d'argent, interactions avec l'administration ou gestion de l'énergie. Loin d'être des projets périphériques, ces initiatives contribuent à définir le rôle de l'Afrique dans l'une des transitions technologiques les plus déterminantes de l'histoire humaine.
Pourtant, une tension structurelle marque ce moment charnière. L'IA ne se déploie pas grâce à une percée unique, mais par sa diffusion : le travail concret consistant à intégrer l'IA au sein d'institutions réelles et de structures de responsabilité solides, afin qu'elle soit fiable, digne de confiance et durable. Or, cette diffusion est impossible sans un socle de confiance.
À travers le continent africain, les startups spécialisées dans l'IA traitent des volumes croissants de données sensibles et exploitent des infrastructures numériques critiques. Elles sont confrontées à des dilemmes complexes, alors même que l'expertise, les outils et le soutien institutionnel en cybersécurité nécessaires pour protéger leurs réalisations ne sont pas adaptés à leurs besoins. Ce n'est pas un manque d'ambition, mais une lacune dans l'écosystème. Si cette faille n'est pas comblée de manière volontariste, les conséquences seront lourdes : failles de sécurité érodant la confiance du public, échecs institutionnels freinant l'adoption, et dégradation progressive de la confiance des utilisateurs face à des systèmes aux comportements imprévisibles et sans mécanisme de responsabilité clair.
La cybersécurité se situe précisément à cet endroit. Et y remédier n'est pas un simple défi technique, c'est un enjeu de marché — un domaine pour lequel personne n'a encore bâti l'infrastructure nécessaire à l'échelle du continent africain.
À mesure que les écosystèmes africains d'IA se développent, la cybersécurité devient un pilier fondamental — et non optionnel — pour bâtir la confiance, la résilience, l'adoption et la viabilité à long terme.
Réfléchissez à ce que représente l'adoption de l'IA sur le terrain, dans les contextes où évoluent les startups africaines.
Une startup de la health-tech à Nairobi déploie un outil de diagnostic par IA dans un réseau de cliniques rurales. Le système traite les dossiers des patients, fournit des recommandations de traitement et fonctionne malgré une connectivité intermittente. Cela fonctionne. L'adoption progresse. Puis, une faille de sécurité expose les données sensibles de milliers de personnes qui ignoraient que leurs dossiers étaient traités par un algorithme. À l'échelle mondiale, cette fuite n'est pas catastrophique. Mais dans un contexte où la confiance envers les institutions est difficile à gagner et facile à perdre, il n'est pas nécessaire qu'elle soit catastrophique pour être fatale. La startup perd ses partenariats avec les cliniques. Les cliniques perdent la confiance de leurs patients. Et la prochaine application de santé basée sur l'IA qui tentera de pénétrer ce marché devra composer avec la méfiance persistante laissée dans le sillage de cet incident.
Suite à l'appel à manifestations d'intérêt, le PNUD s'efforcera de faciliter les conversations avec les organisations participantes autour des questions d'apprentissage. Les organisations participantes auront la possibilité d'accéder à des réseaux, à un financement potentiel pour des projets pilotes et à d'autres ressources que cette phase de conception conjointe réunira par l'intermédiaire de partenaires intéressés.
The AI Hub for Sustainable Development has been working with UNDP’s AI Trust and Safety Re-imagination Programme on a framework that reframes how we think about safe AI adoption. The insight that matters most for this programme is this: trust and safety are not qualities of a model. They are properties of a system – of the institutional environment in which a model operates, the accountability structures around it, the escalation pathways when something goes wrong, and the human capacity to monitor, interpret, and intervene.
Cybersecurity is the foundation of all of this. It is not a layer added after the pathway is designed. It is the condition that makes the pathway viable. For African AI startups, this means that cyber resilience is not a compliance requirement imposed from outside. It is the internal capability that determines whether their systems can be trusted. Startups that build this capability early will not just be more secure. They will be easier to adopt, more attractive to investors, and better positioned to influence the governance discussions that will shape their operating environment.
About the Cyber4Africa Programme
The AI Hub for Sustainable Development was established on a specific premise: that African innovators should not merely participate in the AI revolution but be central to building it. The AI Hub’s programmes on compute access, talent development, infrastructure, and standards addresses specific constraints that holds African AI startups back.
Cyber resilience is the next frontier, and the AI Hub is not approaching it alone.
The Cyber4Africa Programme brings together Cyber 4.0's deep expertise in cybersecurity capacity building, Cisco Kenya's technical infrastructure and regional reach, and the AI Hub's ecosystem position at the intersection of AI development, policy, and institutional trust. Together, these partners are building something that does not yet exist at scale in Africa: a practical, startup-oriented cyber resilience model that is embedded into the AI development lifecycle from the beginning.
The programme is designed around a conviction that the most important interventions are upstream. Not auditing systems after vulnerabilities emerge, but equipping founders and technical teams to build securely from day one. It seeks to embed security-by-design into the innovation process so that trust is earned before it is tested. It also seeks to build a community of practice among African AI founders who understand that their collective security posture shapes the credibility of the entire ecosystem.
This is also, explicitly, a jobs agenda. The same communities most exposed to AI risk are best positioned to become its monitors, auditors, and assurance professionals, if the pathway into that profession is built deliberately. The Cyber4Africa Programme creates that pathway. It trains not just for technical competence but for the professional infrastructure of AI safety. These are the skills the African AI ecosystem needs. They are also the skills the global AI safety market is beginning to pay for.
Security-by-design is not a constraint on innovation. It is the condition that makes African AI innovation sustainable, scalable, and trusted — by users, by partners, and by the world.
Sectoral Focus
The programme targets African AI startups in five sectors where cybersecurity exposure is most acute, and the consequences of failure are most severe:
Healthcare and Digital Health — patient records, diagnostics, and health infrastructure where security failures can compromise care and erode public trust.
Fintech and Financial Services — the most targeted sector globally, where cyber incidents undermine both financial inclusion and consumer confidence.
Govtech and Civic Technology — digital identity and e-government platforms where vulnerabilities directly compromise institutional legitimacy.
Energy and Climate Technology — AI-enabled infrastructure where digital vulnerabilities carry physical and national consequences.
AI Infrastructure and Data Platforms — the foundational layer where security failures cascade across entire networks of users, developers, and dependent systems.
Programme Approach
The programme runs for six months and is structured around three interconnected workstreams:
Cyber Resilience and Awareness: Tailored advisory support to help startups assess their cybersecurity posture, strengthen organisational awareness, and align with international security standards. All participating startups receive a Cybersecurity Light Assessment at the start of the programme.
Training and Capacity Building: Specialized training delivered remotely and in person, including webinars on AI security essentials, a cybersecurity starter mini-course, and access to Cisco's Cybersecurity Training Centre in Nairobi for hands-on workshops.
Technical and Operational Support: Hands-on support to help startups implement cybersecurity controls, strengthen incident response capabilities, and improve operational resilience through expert engagement, technical office hours, and AI Security Quick Scan
Who Should Apply
The Cyber4Africa Programme is open to African AI startups that:
- Have an AI-enabled product, platform, or service at MVP stage or beyond
- Operate in environments with meaningful cybersecurity exposure
- Handle sensitive personal, financial, or operational data
- Are building in one or more of the programme's five priority sectors
- Are Africa-based or building solutions for African markets
- Have the capacity to actively participate throughout the six-month programme
Why Apply
Participants in the Cyber4Africa Programme will:
- Receive a structured cybersecurity assessment and a prioritised action plan tailored to their technology stack and sector context
- Build the technical and operational security capability needed to protect their product, their users' data, and their institutional partnerships
- Access expert-led training, remotely and in-person, from Cyber 4.0's specialist network and Cisco's Nairobi Training Centre
- Strengthen their position with institutional partners, investors, and regulators who increasingly require evidence of security practices before procurement or collaboration
- Join a community of African AI founders committed to building with security and trust at the centre
Applications are now closed.
Thank you to everyone who applied to the Cyber4Africa Programme.
The inaugural cohort has now been selected.
