South Africa’s unemployment rate has climbed to 33.6%, highlighting the depth of the country’s jobs crisis and raising a more complicated question: what role, if any, will artificial intelligence play in a labour market already struggling to create enough work?
The latest Quarterly Labour Force Survey from Statistics South Africa showed that the official unemployment rate increased from 32.7% in the first quarter of 2026 to 33.6% in the second quarter.
The figures underline a labour market that continues to struggle to absorb new entrants, even as businesses and governments around the world look to AI to improve productivity and create new types of work.
However, Dr Lerato Ntuli, economist at Anchor Capital, believes AI is unlikely to have a significant impact on South Africa’s unemployment rate in the short term.
“South Africa's unemployment challenge remains largely structural, driven by infrastructure constraints, weak investment and slow economic growth, all of which limit job creation and the adoption of AI technologies,” Ntuli said.
“In the near term, AI is therefore unlikely to have a significant impact on reducing unemployment,” she added.
SA’s jobs crisis is bigger than AI
South Africa’s unemployment rate has remained above 30% since 2020, while economic growth has remained too weak to generate sufficient employment opportunities.
The country’s annual GDP growth has been below 2% since 2023, Ntuli said.
The weakness in economic growth is compounded by uncertainty in the global economy, including geopolitical tensions, US trade and tariff policies and the ongoing war between Russia and Ukraine.
These conditions make it more difficult for businesses to commit to significant investment and expansion, both of which are important sources of job creation.
The Congress of South African Trade Unions (Cosatu) described the latest figures as deeply concerning and called for urgent action to tackle unemployment.
Cosatu said the expanded unemployment rate had increased to 43.8%, with job losses recorded in community and social services, mining, agriculture and manufacturing.
The union has called for a major stimulus package, greater support for small businesses and industrial sectors, expanded public employment programmes and increased investment in infrastructure.
AI could create jobs, but not enough yet
AI is often discussed primarily in terms of its potential to automate existing jobs. But technology analyst and writer Wesley Diphoko believes the technology will also create new employment opportunities.
He pointed to a development in South Africa’s financial services industry in March 2026, when a major financial services company appointed its Group Chief Technology and Information Officer as its first Group Chief AI Officer on the Group Executive Committee.
The appointment demonstrates how AI is increasingly becoming a strategic priority at executive level.
Diphoko said this is one example of how AI will create new roles as companies build the capabilities needed to manage and implement the technology.
However, he believes its overall impact on unemployment remains limited at this stage.
“In terms of unemployment it seems there's limited impact influenced by AI,” Diphoko said.
The distinction is important for South Africa.
AI may create demand for specialised roles in areas such as data, technology, cybersecurity, AI governance and business transformation. But these opportunities are unlikely, at least for now, to be large enough to make a meaningful dent in an unemployment rate above 33%.
AI could reshape jobs rather than replace them
Ntuli believes the longer-term picture could be more positive.
AI could improve productivity and competitiveness in industries such as financial services while changing the nature of existing jobs.
“Over the longer term, AI could improve productivity and competitiveness in sectors such as finance, while reshaping jobs rather than replacing them outright,” Ntuli said.
However, she warned that South Africa cannot assume that the benefits of AI will automatically reach workers across the economy.
“Realising these benefits will require investment in electricity, digital infrastructure, and skills development to ensure AI adoption is broad based and inclusive rather than exacerbating existing inequalities,” she said.
This could become one of the country’s biggest challenges as businesses accelerate their adoption of AI.
Companies with access to reliable electricity, high-speed connectivity, skilled employees and investment capital will be better positioned to implement AI.
Smaller businesses and workers without access to these resources risk being left behind.
Infrastructure remains a constraint
For South Africa, the AI debate therefore extends beyond the technology itself.
Reliable electricity, functioning transport infrastructure, affordable connectivity and skills development all affect businesses’ ability to invest, expand and adopt new technologies.
Ntuli said South Africa’s structural challenges are currently limiting both employment creation and the adoption of AI.
This means that encouraging companies to invest in AI alone is unlikely to transform the country’s labour market.
The technology could improve productivity, but the effect on employment will depend in part on whether those productivity gains translate into stronger investment, business expansion and new economic activity.
The business decision
For South African companies, the question is therefore not simply whether AI can replace workers or create new jobs. It is what businesses do with the productivity gains that the technology creates.
A company that uses AI to perform work more efficiently may become more competitive and reduce costs. But the wider economic benefit is different if those gains are used to expand production, enter new markets, develop new products and create new opportunities elsewhere in the business.
That distinction matters in an economy where unemployment is already above 30%.
South Africa needs businesses to become more productive, but it also needs that productivity to translate into investment and economic growth.
AI is arriving in an economy that desperately needs productivity and desperately needs employment. The business decisions made over the next few years will determine whether those two needs reinforce each other — or pull in opposite directions.
Economic growth remains central
Cosatu has also warned that South Africa cannot continue to normalise economic growth of around 1% alongside extremely high unemployment, poverty and inequality.
The union has called for measures to reduce electricity costs, restore rail and port capacity, invest in economic infrastructure and tackle crime and corruption.
It also wants greater access to affordable capital for small businesses and industrial sectors.
The latest unemployment figures therefore point to a problem that predates the arrival of AI. South Africa’s ability to create jobs remains closely tied to economic growth, investment and the capacity of businesses to expand.
For now, the evidence suggests AI is unlikely to materially change the country’s unemployment rate. Its longer-term effect will depend on how quickly businesses adopt the technology, whether productivity gains translate into new investment and economic activity, and whether workers have the skills and infrastructure needed to participate in that transition.
ashley.lechman@nationalmg.co.za