AI Job Squeeze Threatens 3.3 Million South African Workers
Africa · Southern
Key Facts
—Automation, not elimination. Generative AI is taking over routine customer-service tasks, but humans still handle complex and sensitive cases.
—Entry-level roles hit hardest. Junior positions in customer service, data entry, and administrative work face the greatest exposure to AI displacement.
—3.3 million jobs at risk by 2030. McKinsey estimates digitisation and automation could displace millions of South African workers while creating 1.8 million new roles.
—Hiring slowdown, not collapse. PwC data shows gentler job-posting growth in AI-exposed sectors rather than sharp outright declines.
—Infrastructure contest. AI is also a geopolitical race over cloud, data, and compute ownership, with African states at risk of becoming buyers rather than builders.
An AI job squeeze is reshaping South Africa’s business-process outsourcing sector, as routine customer-service tasks are automated and entry-level hiring slows in a country where first-rung employment is already desperately scarce.
The quiet restructuring of Johannesburg’s call centres
Across Johannesburg’s sprawling business-process outsourcing (BPO) hubs, generative AI tools are now handling tasks that once filled entire shifts. Live-chat triage, frequently asked questions, password resets, and basic account queries are increasingly managed by software rather than people.
Industry analysts say the change is not yet a wholesale replacement of human workers. Complex, emotional, and compliance-sensitive interactions still require agents, but the volume of routine work feeding entry-level positions is shrinking fast.
PwC South Africa’s 2026 AI Jobs Barometer captures the shift in numbers. The share of job postings requiring AI skills nearly doubled to 3%, while overall labour-market conditions weakened across exposure groups.
Why the AI job squeeze hits South Africa harder
South Africa enters this transition with one of the world’s highest unemployment rates. Call-centre and BPO work has long served as a critical first rung for young people entering the formal economy, particularly in Gauteng and the Western Cape.
A Western Cape government report warns that automation is expected to replace routine tasks in administrative fields and may hit lower-skilled workers hardest. Retail and administrative roles are among the most vulnerable categories identified.
Old Mutual Investments has flagged a deeper social risk. If the first rung of the employment ladder is undermined, social mobility becomes harder and wealth concentration could increase, especially if advanced AI systems are owned by a small number of global firms and investors.
The numbers behind the shift
McKinsey’s South Africa future-of-work research offers the starkest projection. Digitisation, machine learning, and automation could create 1.8 million new jobs through productivity gains while displacing 3.3 million existing jobs by 2030.
More than 500,000 of those displaced positions could come from manufacturing alone. Academic studies from South African universities find AI is associated with weaker low-skilled employment over the long run, even where some skilled jobs gain from adoption.
Yet the picture is not uniformly bleak. PwC’s 2025 barometer describes gentler growth rather than sharp declines in AI-exposed sectors, and a Yale Budget Lab assessment found no discernible broad labour-market disruption from AI globally since ChatGPT’s release.
Who owns the stack: the geopolitics of AI automation
The deeper story is not simply about jobs lost to software. AI is a contest over compute, cloud infrastructure, data, chips, and model ownership, and African countries risk becoming buyers of foreign AI systems rather than owners of productive capacity.
Global technology companies, cloud providers, and governments are competing for access to data, computing power, and new markets across the continent. This makes infrastructure control a strategic issue, not just a business one, as detailed in our pillar series Africa: The New Scramble.
The Brookings Institution has argued Africa should sequence rather than rush into AI adoption. Premature automation could displace workers in sectors such as logistics and regional call centres before alternative employment emerges, creating bargaining space for states but also raising dependence risks.
What the AI job squeeze means for investors and the BPO sector
For international investors and BPO operators, the equation is shifting. South Africa has long marketed itself as a competitive English-language outsourcing destination with a favourable time zone and a large, young workforce.
The question now is whether firms can use AI to boost productivity and retain contracts, or whether labour-intensive work will disappear faster than workers can be moved into higher-value roles. The World Economic Forum projects 170 million new jobs created and 92 million displaced globally by 2030, but the net effect for South Africa depends on reskilling speed and infrastructure investment.
A World Bank blog offers a qualified counterpoint, arguing that innovation and high-tech sectors can generate employment in South Africa overall even as automation threatens some low-skilled work. The outcome hinges on policy choices made now.
The BRICS and South-South dimension
South Africa’s AI transition unfolds within the BRICS framework, where digital cooperation and technology transfer are increasingly central to bloc discussions. Fellow BRICS members China and India are major AI developers, while Brazil faces parallel questions about automation in its own large service sector.
For Latin American readers, the South African experience offers a preview of pressures that are already visible in Brazilian and Mexican call-centre markets. The same global platforms, cloud vendors, and AI models are being deployed across the Global South, often with limited local ownership of the underlying technology.
The bargaining power of African and Latin American states will depend on whether they can negotiate infrastructure investment, data sovereignty, and reskilling programmes as part of AI adoption, rather than simply absorbing job displacement as an externality of foreign-owned platforms.
Connected Coverage
Frequently Asked Questions
Is AI eliminating call-centre jobs in South Africa entirely?
No. Generative AI is automating routine tasks such as live-chat triage and FAQs, but humans are still retained for complex, emotional, and compliance-sensitive cases. The near-term effect is job restructuring, slower hiring, and reduced entry-level intake rather than mass replacement.
How many South African jobs could be displaced by automation by 2030?
McKinsey estimates that digitisation and automation could displace 3.3 million existing jobs in South Africa by 2030 while creating 1.8 million new roles. More than 500,000 of those displaced positions could come from manufacturing alone, with entry-level service roles also heavily exposed.
Why does AI automation matter more in South Africa than in richer economies?
South Africa has exceptionally high unemployment, and call-centre work serves as a vital first-rung job for young people entering the formal economy. If AI removes that rung without creating alternative pathways, inequality could deepen and social mobility could stall, according to Old Mutual Investments research.
Sources
Sources: McKinsey; PwC.
LatAm Markets: Live Signals → — real-time movers, turnover leaders and FX across Latin America.
Read More from The Rio Times