By 2026, South African business leaders are shifting their focus from whether Artificial Intelligence matters to how quickly it can be applied responsibly, transforming not just jobs but the very fabric of work itself.
Image: AI / Sora
In 2026, South African business leaders are no longer asking whether Artificial Intelligence (AI) will matter: they are asking how fast they can apply it responsibly.
You can see the shift in boardrooms and operating models. AI is moving from a side project to a core capability, visible across finance, customer operations, marketing, risk, legal and HR, anywhere there is a repeatable decision, a workflow, or large volumes of information to sift through.
The biggest shift is not the technology itself, but what follows.
Operating models are changing. Teams will be built differently, performance will be measured differently, and judgement will matter more, not less.
Executives are grappling with a paradox.
AI can make organisations leaner in routine work.
At the same time, it raises the premium on skills that are anything but routine. Problem framing, prioritisation, commercial judgement, ethics, and the ability to work confidently with data all become more valuable, not less.
That is why debates about “AI and jobs” often miss the point.
The more immediate impact is not mass job loss, but a change in the structure of jobs. Across sectors, tasks are being unbundled and re-bundled, workflows redesigned, and decision-making pushed into smaller teams equipped with better tools.
In banking, the first wave is less about removing jobs and more about redesigning frontline and back-office work.
AI is helping summarise interactions, automate reconciliations, and surface exceptions that require human judgement.
In retail, it is reducing friction in merchandising and supply chain planning, while giving store and contact-centre staff faster answers and clearer escalation paths. In telecoms, it is improving network and customer-care triage, again changing the shape of roles more than simply shrinking them.
AI-driven efficiency will reduce the amount of time, and sometimes the number of people, needed for certain types of administrative and transactional work.
But the more important shift is qualitative.
As automation and AI assistants take over parts of a process, such as drafting, first-pass analysis and case triage, human work moves up the value chain towards interpretation, stakeholder engagement, and accountable decisions.
For South Africa, with high unemployment, uneven schooling outcomes and a stretched skills pipeline, this redesign of work matters as much as any net change in headcount.
If businesses treat AI purely as a cost-cutting lever, the social and reputational consequences will be real. If instead we treat it as a productivity lever and build capability alongside it, we have an opportunity to grow output per person and widen pathways into higher-value work.
It would also be a mistake to assume that more AI means less human contribution.
In many functions, AI raises the bar for what good looks like, clearer thinking, stronger writing, better analysis, and the confidence to challenge an output rather than accept it at face value.
Some roles will become more specialised and technical. Others will become broader, with fewer hand-offs and more end-to-end accountability. Either way, workforce planning has to shift from asking “how many people do we need?” to asking “what work should humans do, what can machines do well, and how do we redesign the system so quality improves?”
This is where business leaders need to get specific.
A practical starting point is a skills-and-tasks view of each function, identifying where AI can remove friction.
From there, investment should flow to what remains scarce, data literacy, process redesign, risk awareness and change leadership. Some targeted hiring will still be needed in areas such as data engineering, cybersecurity and model risk. But the bigger prize is systematic reskilling so existing teams can work confidently alongside AI.
AI can lift productivity dramatically, but only when people bring context, accountability and the ability to test an answer against reality.
This is also why critical thinking should be seen as a national competitiveness issue, not a nice-to-have. It needs to start at school, with the ability to reason, read, write and work with evidence. It should be reinforced at university through disciplined research and debate.
And it must carry into the workplace as a habit, asking better questions, verifying sources, and knowing when to escalate uncertainty.
On the technology side, many South African organisations are moving from experimentation to execution.
Early AI investments are starting to show measurable value, faster turnaround times, improved customer service, lower error rates, better forecasting and smarter risk detection. In banking, that can mean quicker onboarding and cleaner reconciliations.
In retail, better demand signals. In telecoms, sharper triage in customer care and network operations. That is why budgets are shifting from “innovation” to “run the business”.
Over the next two years, expect acceleration.
The winners will be firms that embed AI into a small number of practical, high-impact use cases, keep improving them, and tie them to clear operational ownership. In most cases, the real differentiator is not the model itself, but data quality, process redesign and adoption.
As AI becomes embedded in core processes, attention is also turning to practical risk questions. What data is being used? Who is accountable for outcomes? What information should never be shared with external tools? The aim is not to slow adoption, but to scale it with clear rules, sensible controls and people who know when to challenge an output.
The thread running through all of this, it is that the South African enterprise that succeeds will hold three ideas together. Use AI to simplify and speed up work. Redesign roles rather than simply reduce them. Build the judgement that keeps decisions anchored in reality.
The age of AI execution has arrived.
It calls for leaner processes, sharper skills and more serious oversight, not panic.
If we get the balance right, AI can become a powerful lever for productivity and competitiveness in South Africa.
Ajen Sita, CEO, EY Africa
Ajen Sita, CEO, EY Africa.
Image: Supplied.
Follow Business Report on Facebook, X and on LinkedIn for the latest Business and tech news.