The Zimbabwean-born founder of Econet and Cassava Technologies has announced plans to build artificial intelligence computing capacity on African soil. The question is whether the continent can afford to run it.

Strive Masiyiwa built a telecoms and fibre group across Africa after a long legal fight to obtain a licence in Zimbabwe. His current vehicle, Cassava Technologies, spans fibre, data centres and cloud services. Its most publicised move has been a partnership to deploy advanced AI computing capacity in African data centres, starting in South Africa and extending to other markets.

Why it is significant

Almost all AI computation used in Africa happens abroad. That creates three dependencies: cost, since capacity is billed in dollars; latency and data residency, which matter for regulated sectors; and sovereignty, since the rules under which African data is processed are set elsewhere.

Building capacity locally addresses all three, in principle. In practice it requires three things Africa struggles with: enormous quantities of reliable electricity, cooling, and customers able to pay.

The questions for the profile

• How much capacity has been installed, where, and who is using it?
• What does it cost per hour compared with the same capacity bought from a global provider?
• How is the power supplied, and at what carbon cost?
• What is the plan for Central Africa, if any?

The wider lesson

Masiyiwa’s career is a case study in infrastructure entrepreneurship: fight for the licence, build the network, then sell services over it. The AI story follows the same pattern. Whether it works will tell us whether the continent’s digital economy can own its own means of production, or whether it will rent them indefinitely.