Leadership · 2 min read
AI and human leadership in Africa
The machine will do the analysis. The leader still has to decide, carry the consequence and keep the trust of the people. On this continent, that division of labour is an advantage.
B57 Group
There is a lazy version of the future in which AI replaces leaders. It is wrong, and it is dangerous, because it lets leaders off the hook. The real shift is narrower and more demanding. AI takes over the part of leadership that was always about information: gathering it, summarising it, modelling it, forecasting from it. What is left is the part that was always about people.
What the machine takes
A permanent secretary used to wait a fortnight for a briefing. Now the briefing can be assembled in minutes from every report the ministry has ever filed. Budget scenarios that took a team a month can be run before lunch. Correspondence can be triaged, drafted and translated on arrival.
All of that is analysis, and analysis was never the point of leadership. It was the cost of it. Removing the cost does not remove the job. It exposes it.
What stays human
Three things do not move to the machine, and no serious leader should want them to.
Judgement under uncertainty. A model can tell you the probability that a road will flood. It cannot tell you whether to spend the money on the road or the clinic. That is a choice about what a community values, and someone has to make it in public.
Legitimacy. People follow leaders they recognise as theirs. A decision announced by a system has no author and therefore no authority. A decision announced by a person who will answer for it does.
Responsibility. When the decision is wrong, a human must own it. A model has no reputation to lose. A leader does, and that is exactly why leaders can be trusted with power.
Analysis was never the point of leadership. It was the cost of it.
Why Africa is well placed
Africa’s leadership traditions are consultative at their root. A chief sits with elders. A decision is argued through before it is announced. Consensus is a method, not a weakness. That is precisely the posture the AI age demands: take in many inputs, deliberate openly, decide with authority, and answer for the result.
There is a practical advantage too. Much of the continent is not carrying forty years of legacy systems and the institutional habits that come with them. A ministry that never fully computerised can move straight to AI-native operations, the way the continent moved straight to mobile money. The leaders who understand this will not be catching up. They will be setting the pattern.
The one thing to get right
The danger is not that leaders will use AI. It is that they will outsource judgement to systems built elsewhere, trained on other societies, and answerable to nobody here. A minister who asks a foreign model what Ghana should do has handed over the one part of the job that was theirs.
Sovereign AI is the answer to that. Models trained on our languages, our laws and our markets, running on infrastructure we hold, used by leaders who know they are responsible for the outcome. The machine advises. The African leader decides.