AI boom may leave Africa’s poorest behind – World Bank

Artificial intelligence could help cut poverty and expand the digital economy in Nigeria and other African countries, but the poorest households risk being shut out of its gains, the World Bank has warned.

In its October 2026 Africa Economic Update, the bank said AI could serve as a poverty-reducing platform, but access to affordable internet, devices, electricity and digital skills remains limited.

Early signs are encouraging. Nigeria’s GitHub developer base has grown tenfold since 2020, while Ghana’s has risen nearly eightfold, with registrations accelerating after free AI coding assistants became available.

“Some early signals are encouraging, showing that African countries have become increasingly active in software development since the AI boom,” the report said.

The bank said this hints at what wider access could deliver: a larger digital services sector, stronger regional collaboration and productivity gains in agriculture, education, health and finance.

But it warned of “a fundamental constraint: only a relatively well-off minority can reliably access and afford the technologies through which it is delivered.”

Mobile internet in Sub-Saharan Africa is the least affordable in the world relative to income. A basic data package costs about twice the United Nations’ affordability target of two per cent of average monthly income. An entry-level handset costs the poorest fifth of the population about three-quarters of a month’s income.

But electricity compounds the problem.

“Mobile phone ownership alone does not provide meaningful digital access without reliable electricity,” the report said.

Across 19 African countries with data, only 12 per cent of households in the poorest income quintile have both a phone and a grid connection, compared with 54 per cent in the richest. In many countries, fewer than 10 per cent of adults in the poorest quintile are online, against 60 to 70 per cent or more among the richest in better-connected economies.

The bank said AI’s productivity gains could therefore concentrate among people and businesses already better positioned, leaving farmers, informal businesses, schools, clinics and underserved communities behind.

“The immediate risk for most African economies is not mass displacement of poor workers by AI, but rather that the productivity gains from AI accrue almost entirely to highly educated workers, formal firms, and richer urban households,” it said.

The bank urged countries to move from connecting markets to connecting poor people and places, through investment in connectivity, electricity, digital skills and computing capacity.

“If they do not, it will amplify the advantages of those already connected,” it said.