The AU set the what and the why. Africa needs the how.
The Continental AI Strategy gives every African government a mandate. Turning it into funded programmes, working systems and governed operations is where an institution earns its place.
In 2024 the African Union adopted its Continental Artificial Intelligence Strategy. It settled two questions that had been open for a decade. The what: priorities in talent, data, infrastructure, markets and governance. The why: a case for urgency, endorsed at the highest level of the continent, that opens the door of every ministry.
What the strategy leaves open, deliberately, is the how. How does a national AI strategy become a programme with an owner, a budget and a delivery date? How does a ministry procure an AI system without surrendering its data, its audit rights or its future choices? How does a pilot survive the eighteen months between the demonstration and the budget line?
The delivery gap
This gap is not a failure of ambition. Declarations and headline figures are plentiful. What is scarce is the institutional machinery that turns a mandate into a working system: readiness baselines a finance ministry can fund against, procurement routes a treasury will approve, and operating teams that stay after the consultants leave.
Africa Tech Trust was built for that gap. Its AI practice runs on three business lines. Build & Operate delivers AI systems for governments and institutions — from paid pilot to managed operations at national scale. Enable & Govern builds the governance, safety and institutional capability that make adoption controlled rather than improvised. Own & Scale works towards sovereign AI, data and platforms — African-controlled, reusable across countries, licensed as models, datasets and intellectual property.
Every assessment ends in a roadmap
The first engagement is never a report. A Sovereign AI Readiness & Implementation Assessment takes six to eight weeks and ends with a readiness baseline, one prioritised pilot scoped and costed, a procurement route and a twelve-month roadmap. The pilot-to-production programme that follows takes twelve to sixteen weeks and ends with a working system on the institution’s own data, governance controls, a production plan and a managed-service proposal.
- Assess — a baseline and a prioritised roadmap
- Pilot — a working system with governance controls, in weeks
- Implement — production deployment, integration and evaluation
- Operate — managed AI operations, monitored and governed
The pathway is deliberately linear. Each stage produces something the client can use even if the next stage never happens; each stage is also the natural entry to the next.
Own accountability, not every resource
ATT holds the client relationship, the architecture, the governance and the quality assurance. African partners, startups and universities deliver engineering, models, data preparation and cloud alongside it. Capital-intensive infrastructure — data centres, frontier models — is built only through a dedicated entity with contracted demand. That discipline keeps the institution honest and the client in control.
The AU document is the mandate. The delivery gap is where an institution earns its place.
Governments, agencies and institutions that want to move from strategy to funded programmes can request a briefing. It takes an hour, and it ends with a next step.
The ATT Brief — the monthly briefing, by email