40% of Africa’s Outsourcing Tasks. Not 40% of the Jobs.
A repeated headline says AI could replace 40% of jobs in African outsourcing by 2030. The report it comes from says tasks, and projects the sector will grow.
If you have read anything about artificial intelligence and African outsourcing in the last year, you have probably seen a version of this headline: 40% of jobs at risk by 2030.
The report it comes from does not say that. It says 40% of tasks.
What the report actually found
The study, published in April 2025 by Caribou Digital and Genesis Analytics in partnership with the Mastercard Foundation, covered South Africa, Rwanda, Kenya, Nigeria and Uganda. Researchers identified 130 distinct tasks across 36 key job roles and scored each for automation potential over a five-year horizon.
Botswana is not one of the five countries. Nothing in the report is a measurement of Botswana, and we have not found an equivalent study that is. What it offers a reader here is a method and a set of task-level findings, not a local forecast.
Over 40% of current tasks scored as susceptible. About 10% were judged fully resilient. That leaves roughly half in between, partially automatable, which is the definition of a job changing rather than disappearing.
Finance and accounting scored highest at 44%. Customer experience, which accounts for 44% of sector employment, also came in around 40%.
The part the headlines dropped
The same report projects the sector adding about 1.8 million direct jobs by 2030, with customer experience growing at 24% compound annually. It is not forecasting contraction.
The authors are also unusually direct about the limits of their own method. The baseline scoring, they write, "does not factor in changes in overall demand for work", and of the headline number itself they say that assuming no change in AI capability or market response "is, of course, unrealistic". This is scenario analysis, and it says so.
Who is actually exposed
The distributional findings deserve more attention than the headline. Work done by women scored about 10% more susceptible to automation than work done by men, because women are concentrated in junior roles and in customer experience. Customer experience is 72% youth and 56% women.
So the exposure is real, and it falls hardest on young women in entry-level roles. That is a more useful thing to know than a wrong number about total jobs.
What the measured evidence says
There is one study that observed rather than projected, and it studied a workforce that was 89% offshore. It found productivity gains, not displacement. We cover it in what AI actually did to offshore support work.
The International Labour Organization’s exposure index adds useful context: about a quarter of workers globally are in occupations with some exposure, but that runs at 11% of employment in low-income countries against 34% in high-income ones, because poorer economies have fewer of the office roles these tools target. Outsourcing is the exception, since it exists precisely to host those roles. The tension between those two facts is the actual story, and it is more interesting than the headline. See also the graduate employment question and the digital marketing guide.
Frequently asked
Will AI replace 40% of jobs in African outsourcing?
No, and the report behind that headline does not claim it. It found over 40% of current tasks susceptible to automation by 2030, with only about 10% fully resilient, which leaves roughly half partially automatable. The same report projects the sector adding around 1.8 million jobs by 2030.
Which outsourcing roles are most exposed to AI?
Finance and accounting scored highest at 44% of tasks, with customer experience close behind at around 40%. Because women and young workers are concentrated in junior and customer-experience roles, work done by women scored roughly 10% more susceptible than work done by men.
Is Botswana more exposed to AI job losses than richer countries?
At the whole-economy level, less. The ILO estimates 11% of employment in low-income countries has some generative AI exposure against 34% in high-income countries, because poorer economies have fewer office-based roles. The outsourcing sector is the specific exception, since it exists to host exactly those roles.