Innovation & Tech

Kenya topped ChatGPT use; Microsoft's data puts South Africa on top

DataReportal's July 2025 survey put Kenya first in ChatGPT use at 42.1% of internet users. Microsoft's 2026 data ranks South Africa first at 23.9%. The gap is mostly the denominator.

By Olivia Hart

5 min read

Who is actually using AI in Africa? It depends on whom you count
Who is actually using AI in Africa? It depends on whom you countLalmch / pixabay

What's News

  • 42.1% of Kenyan internet users aged 16+ reported monthly ChatGPT use in DataReportal's July 2025 survey, ahead of South Africa at 34.3% and Nigeria at 27.9%
  • Microsoft's AI Diffusion Report put second-quarter 2026 generative AI use at 23.9% of South Africans aged 15-64, 10.4% of Nigerians and 8.9% of Kenyans
  • Digital 2027 reports put internet penetration at 40.6% in Kenya, 47.5% in Nigeria and 79.5% in South Africa
  • 97.3% of Kenyan internet users and 96.6% of Nigerians reported using some form of AI in the previous month, per DataReportal's broader survey
  • Microsoft's mobile usage adjustment rests on assumptions that are, in the researchers' words, "difficult to validate"

Kenya led the world in reported ChatGPT use last year: 42.1% of Kenyan internet users aged 16 and above said they had used ChatGPT in the previous month, according to DataReportal's July 2025 survey. South Africa ranked second at 34.3%, Nigeria third at 27.9%.

Microsoft's newer data reverses that order. In the first quarter of 2026, the company's AI Diffusion Report estimated that 23.1% of South Africans aged 15 to 64 had used a generative AI tool, against 10.1% of Nigerians and 8.7% of Kenyans. By the second quarter, South Africa reached 23.9%, Nigeria 10.4% and Kenya 8.9%.

Both sets of numbers circulate as "AI adoption in Africa." They measure different populations, and neither answers the question investors and ministries need answered: how many people have found a use for AI that brings them back.

Why do two rankings disagree?

DataReportal's ChatGPT figures come from GlobalWebIndex, which surveys internet users. Take two countries with 100 adults each. In the first, 40 are online and half of them use ChatGPT, giving 20 users. In the second, 80 are online and a third use it, giving 27.

The first country leads among internet users. The second has more ChatGPT users across the adult population. The ranking flips on connectivity alone.

That connectivity gap is substantial. DataReportal's Digital 2027 country reports put internet penetration at:

  • Kenya: 40.6%
  • Nigeria: 47.5%
  • South Africa: 79.5%

Those figures cover the whole population and come from a different period than the original ChatGPT survey. Multiplying them by the ChatGPT shares would mix age groups and timeframes and yield a misleading national estimate.

In short, the rankings diverge because the denominators diverge, and neither survey publishes a single comparable national adoption rate.

What happens when the denominator widens?

DataReportal's more recent global AI report widened the question from ChatGPT to AI broadly. It found that 97.3% of Kenyan internet users aged 16 and above reported using some form of AI in the previous month. Nigeria came in at 96.6%.

The report itself warns against the obvious misreading. "It's critical to stress that these figures represent the share of Internet users," the report says. The figures cannot be read as evidence that ChatGPT use has doubled, or that almost everyone in either country uses AI.

Age skews the picture further. Across the markets DataReportal surveyed, 93.3% of internet users aged 16 to 24 said they had used AI in the previous month, compared with 52.2% of those aged 65 and above.

That gap matters in Africa because a younger online population lifts any adoption rate on its own. The published rankings leave open how much of Kenya's lead reflects the ages of those surveyed and how much reflects behaviour among people of the same age.

The broader AI numbers are not a doubling of ChatGPT use, and the age mix of each country's online population remains unquantified.

What can Microsoft actually see?

Microsoft starts from a wider base: everyone aged 15 to 64, whether they work, study or have internet access. It also counts activity across generative AI products. Its own technical paper limits the claim. The data it observes directly are "primarily originating from users on Windows desktop (PC/tablet) platforms," the paper says.

People can opt out of sending diagnostic data, and Microsoft uses other sources to estimate usage beyond Windows and on mobile devices. "These assumptions are difficult to validate," the researchers write of the mobile adjustment.

That limitation lands where it hurts most. For many users, a first encounter with generative AI comes through a chatbot on a phone. A survey captures that experience if the respondent remembers it and reports it; a model anchored to desktop activity has to estimate what it misses.

Microsoft also pools markets with limited data into regional averages. Nigeria's second-quarter 10.4% estimate is shared with Ghana, Benin and Togo. The table alone offers little basis for treating Nigeria's 1.5 percentage-point lead over Kenya as a precise measure of how differently people in the two countries use AI. South Africa's much wider lead is the more striking result.

Microsoft's figures cover more people but rest on an unvalidated mobile adjustment, and Nigeria's number is a regional average rather than a national one.

Does adoption translate into revenue?

For a startup selling an AI writing tool to small businesses, familiarity with ChatGPT makes the pitch easier. Pricing calls for different evidence: how often do those businesses need help writing, how much do they already get from a free chatbot, and would a tool working in the languages they use with customers justify a subscription.

A high adoption rate among internet users tells a founder where to look for customers. It says little about what those customers will pay.

For a government, a wrong denominator carries a different cost. A ministry that puts an AI assistant on its website to explain how to apply for a permit might read Kenya's ChatGPT share as evidence that connected applicants will find it familiar. The ministry still needs to know how many applicants can reach the site, whether they can follow the instructions on a phone, and how often they complete the application. Even Microsoft's wider estimate excludes people aged 65 and over, who may also need the service.

Neither headline figure separates curiosity from habit. A student who asks ChatGPT to summarise one reading and a trader who uses it every week to draft replies to customers both count in a monthly survey. The distinction starts to matter when the trader changes how they handle enquiries because of the tool.

The missing measurement is repeat use, so pricing, public-service design and market sizing all rest on a number nobody has yet asked the right question of.

Where the numbers point next

Investors sizing African AI markets and ministries deploying public AI assistants are working from two incompatible figures, not one adoption rate. The next useful dataset would track whether users return to these tools week after week, and whether the tool changes what they do next.

Source: techcabal.com

Share this article:

More from Olivia Hart

Olivia Hart

Show full bio

Staff writer covering people, launches and day-to-day practice in entrepreneurship, startups and venture funding at Business Bearings.

2,404 articles

Related articles

« Previous articleNext article »