AI Job Atlas
AI & jobs · Country profile

South Africa

Upper-middle income Capital: Pretoria 9 regions
27%
of jobs are in occupations exposed to generative AI. That's about 4.9M people.
MODEL72nd of 183 most exposed World 21%Upper-middle income 21%

Key numbers

Workers in exposed jobsMODEL
4.9M
of 18.1M in work
Highest-exposure jobsMODEL
6.1%
1.1M people
People in workILO
18.1M
45% women
UnemploymentWORLD BANK
32.4%
2025
Youth unemploymentWORLD BANK
59.9%
ages 15–24
GDP per capitaWORLD BANK
$6,598
US$, 2025
Internet usersWORLD BANK
78%
of population
AI preparednessIMF
0.50
68th of 165
PopulationWORLD BANK
64.7M
Tertiary enrolmentWORLD BANK
23%
gross ratio
Vulnerable employmentWORLD BANK
12%
own-account & family workers
Urban populationWORLD BANK
64%

Women and men

MODEL

Based on the real occupation mix of women’s and men’s jobs (ILO).

Women: any exposure 34%
Men: any exposure 21%
Women: highest exposure 9.2%
Men: highest exposure 3.6%

Women are +13.3 pts more exposed, mostly because they hold more clerical and administrative jobs.

How it compares

MODEL
South Africa 27%
Upper-middle income 21%
Sub-Saharan Africa 15%
World 21%

AI preparedness

IMF
South AfricaUpper-middle income average
Digital infrastructure 0.12
Human capital & labour policy 0.12
Innovation & integration 0.11
Regulation & ethics 0.14

IMF AI Preparedness Index (2023). The four parts sum to the overall score of 0.50 (0–1 scale).

Exposure vs preparedness

South Africa is highlighted. The other dots are all countries with both measures.

Sources: ILO modelled estimates 2025 · World Bank WDI · IMF AIPI · Anthropic Economic Index · Wikidata · Natural Earth. How the estimates are made.

Regions of South Africa