Every economy has a first rung. It is where people with little experience enter the labour market, learn how organisations work, develop confidence, and begin building a career. In South Africa, call centres have served as that first rung for hundreds of thousands of young people.
That first rung is beginning to disappear.
The reporting this week on South Africa's business process outsourcing (BPO) sector is worth paying attention to, not because call centres are collapsing, but because they reveal something much bigger about the future of work.
Generative AI is steadily taking over the routine tasks that once filled an entry-level shift. Live chat triage, password resets, basic account queries, and frequently asked questions are increasingly handled by software. PwC South Africa's 2026 AI Jobs Barometer found the share of job postings requiring AI skills nearly doubled to 3 percent this year, while McKinsey estimates automation could reshape millions of South African jobs by 2030. Call centres are simply one of the first places where this transition is becoming visible.
This is not mass elimination. Complex, emotionally sensitive, and compliance-heavy interactions still require human judgment, and they are likely to do so for years to come. What is shrinking is not the industry itself. It is the volume of routine work that once justified hiring someone with little or no experience and training them on the job.
For a country with one of the highest youth unemployment rates in the world, that distinction matters enormously.
Call centres have never been just customer service operations. They have been one of South Africa's most reliable pathways into formal employment. If that pathway narrows, fewer young people gain the experience that allows them to build careers. As Old Mutual Investments recently argued, weakening these entry points risks slowing social mobility and widening inequality.
The problem is bigger than call centres
The greatest disruption may not be that AI replaces jobs. It is that AI is replacing the jobs through which people learned to become professionals.
Routine work has always served a purpose beyond productivity. It gave graduates and school leavers an opportunity to learn how organisations function. They answered customer calls, prepared reports, coordinated projects, tested software, solved problems, and gradually developed judgment through experience.
AI is compressing that apprenticeship.
The routine tasks that once gave people room to learn are disappearing faster than new pathways are emerging.
Call centres make this especially visible because they have long been one of the largest and most measurable entry points into the labour market. Hiring can be tracked. Seats can be counted. The shift is happening in plain sight.
The same transition is unfolding across legal services, finance, marketing, administration, software development, and many other knowledge-intensive industries. It is simply less visible.
There is another dimension to this story. Many of the AI systems replacing routine work are designed, owned, and monetised outside Africa. A customer service agent in Johannesburg is increasingly working alongside, or competing with, software built in San Francisco, London, or elsewhere. The economic value created by automation does not necessarily remain where the jobs are being transformed.
That is why conversations about infrastructure and conversations about human capability cannot be separated. Africa certainly needs more compute, stronger AI ecosystems, and greater technological sovereignty. But even if every data centre the continent required were built tomorrow, the central question would remain unchanged.
What should a young person entering the labour market know how to do when routine work is no longer the starting point?
Buying time is not the same as building readiness
Some researchers, including those at the Brookings Institution, argue that African countries should sequence AI adoption carefully so workers have more time to adapt. That is sensible advice, and governments should think carefully about how to manage the pace of transition.
But sequencing buys time.
It does not build readiness.
A slower rollout of AI in sectors such as BPO might soften the immediate impact, but it does not answer the more important question.
What should a young person leaving school in 2028 know that prepares them for the labour market AI is creating rather than the one AI is replacing?
The old first rung is not coming back.
The high-volume, low-complexity customer service role made economic sense because people were cheaper than software. That economic reality is changing.
What is emerging instead is a different set of opportunities. Organisations will increasingly need people who can resolve complex customer situations, supervise AI systems, audit AI-generated responses, manage exceptions, improve workflows, and combine technical fluency with sound human judgment.
Those opportunities are real.
But they are fewer in number and more demanding than the jobs they replace.
Building the next first rung
This raises a question that extends well beyond call centres.
If yesterday's first rung is disappearing, what becomes tomorrow's?
The answer cannot simply be "learn AI."
Most young people do not need to become machine learning engineers. They need opportunities to develop judgment, communication, adaptability, problem solving, and the ability to work effectively alongside increasingly capable AI systems. They need new forms of apprenticeship that prepare them for work where humans supervise, orchestrate, and improve AI rather than compete with it on routine tasks.
This is why education and workforce development must evolve faster than any single industry.
The young South African who might once have found a stable first job answering customer calls now needs to develop capabilities that AI cannot easily replicate: knowing when to escalate a difficult case, understanding context rather than simply following a script, managing emotionally charged conversations, and making decisions when there is no obvious answer.
None of these capabilities are exotic.
In many ways, they are the qualities that the best call centre professionals have always demonstrated. What has changed is that they can no longer be developed gradually through years of repetitive work. They increasingly need to become the starting point rather than the destination.
South Africa's call centre workforce is not lacking in talent. Many workers have already developed exactly the qualities the emerging economy will reward: resilience, empathy, careful listening, and calm decision making under pressure. What is missing is a deliberate bridge between those capabilities and the new opportunities that are emerging.
Picture what that bridge actually looks like on a call centre floor. An agent who once spent her shift resetting passwords now spends it on the calls an AI system has already flagged as high-risk, a customer threatening to cancel, someone in genuine distress who typed short, tense answers into a chatbot before being escalated to a human. The technical task, resetting a password, required almost no judgment. The new task, calming an angry customer and deciding whether to offer a retention discount or let them go, requires exactly the judgment she already had. It was never absent from her work. It was buried under the routine tasks that used to fill most of her day. The routine tasks are gone. What is left is the part of the job that was always hardest, and always most valuable.
At Mozisha, this is what we describe as the human layer. As AI becomes more capable, uniquely human capabilities become more valuable, not less. The organisations that thrive will not simply be those with access to the best AI. They will be those with people who know how to direct it, challenge it, improve it, and create value alongside it.
Every economy needs a first rung.
For decades, South Africa's call centres played that role for hundreds of thousands of young people.
As AI reshapes routine work, the challenge is no longer preserving yesterday's first rung. It is building tomorrow's before an entire generation reaches for a ladder that is no longer there.