Africa’s AI Divide Is an Instructional-Design Question, Not Just an Infrastructure Problem

Artificial intelligence is often presented as a race for the newest model, the fastest automation or the largest technology budget. That framing misses the decisive question for African education: can the learner actually participate, understand, practise and receive trustworthy support? A system that is technically impressive but inaccessible, ungrounded in the curriculum or unavailable when a learner needs help is not educational transformation.

This matters particularly for theological colleges, church training programmes and distributed adult learning. Our students may be studying after work, sharing devices, managing data costs, moving between urban and rural settings and learning in more than one language. The right response is not to lower academic standards. It is to design learning honestly around the conditions in which people are actually being formed.

A June 2026 preprint by Abayomi Agbeyangi and Jose Lukose examines Africa’s AI divide through infrastructure, accessibility and human capacity. The authors report constrained broadband, high relative data costs, gendered access gaps, limited local language representation and less than one per cent of global data-centre capacity. As a preprint and a continent-wide analysis, it should not be treated as a final description of every country or institution. Its core warning is nevertheless important: infrastructure determines who gets to learn, create and govern with AI.

For curriculum leaders, this turns access into a design responsibility. A course should not assume…