The Strategy That Cannot Say No: Africa's AI Sovereignty Paradox

A national AI strategy is a strange kind of document. It is meant to project sovereignty — to announce that a country has a plan, an ambition, a seat at a table that until recently was set for others. And in 2026, the strategies emerging from Africa's largest economies do project exactly that. Nigeria, Egypt, Kenya and South Africa have each drafted policies that name the problem out loud: their artificial-intelligence futures depend, to an uncomfortable degree, on a handful of American companies — Google, Microsoft, Nvidia and Meta — and they would very much like that to change.

What makes the documents remarkable is not the ambition. It is the candour. These are strategies that, between the lines, concede they cannot yet do the thing they exist to demand. A sovereign AI capability requires compute — vast banks of specialised chips, the data centres to house them, the power to run them — and on every one of those inputs, the continent is starting from almost nothing. The strategy that asserts independence is written on infrastructure it does not own.