In Washington and Brussels, the future of jobs in the age of artificial intelligence is often framed as a story about lawyers using chatbots, programmers sharing code with copilots, and middle managers learning to “prompt engineer” their way through meetings.

In Lagos, Nairobi or Kinshasa, the picture looks very different.

There, the AI economy is already visible in young graduates doing content moderation for global platforms, women working late nights in call centers and data-labeling hubs, and farmers checking prices or weather forecasts on simple smartphones while banks quietly roll out AI-based credit scoring systems they barely understand. Many Africans are not designing the algorithms that will change their work; they are living inside them.

From the Global South, and especially from Africa, the debate about “the future of jobs” is not a futuristic thought experiment. It is an urgent question about whether AI will entrench a global division of labour — with high-value AI design in the North and precarious, semi-automated work in the South — or help rewrite it.

1. A young continent meets an old fear

Africa is the world’s youngest continent. By 2030, one in four young people on the planet will be African. For decades, policymakers have asked: where will the jobs come from?

The World Bank’s Future of Work in Africa report describes a labour market already under strain: high youth unemployment, widespread informality, low productivity and a fragile manufacturing base. It argues that digital technologies could create new pathways for job creation, but only if countries match tech adoption with investments in skills, infrastructure and institutions. World Bank+1

Now generative AI arrives on top of these structural challenges. The worry is simple: if rich countries automate faster, they may “reshore” service work that had started moving to Africa — call centers, back-office processing, basic programming — just as the continent was beginning to benefit from it.

Yet early evidence is more nuanced than the headline fear of “robots stealing African jobs.”

A study by the Center for Global Development points out that, so far, income concentration and job losses directly caused by robots and AI have been limited, and policy choices have mattered more than technology itself in shaping inequality. It warns that automation could slow Africa’s hopes of catching up through labour-intensive manufacturing, but notes that export manufacturing jobs are “not going away yet.” Center For Global Development

In other words, AI does not arrive on a blank canvas. It lands in economies where underemployment is already the norm and where the biggest job killers have often been macroeconomic shocks, political instability or bad policy, not machines.

2. Where African jobs are most at risk

2.1 The fragile promise of the outsourcing boom

Over the last decade, many African governments have bet on business process outsourcing (BPO) and IT-enabled services as a way to absorb young, urban, English-speaking workers. From customer support to basic coding, these jobs promised global wages without migration.

AI now puts that model under pressure.

A detailed task-level analysis of Africa’s BPO and IT-enabled services sector finds that more than 40 percent of tasks across 36 job roles could be automated in the next five years, with finance, accounting and customer service among the most exposed. cariboudigital.net+1

A related study, supported by the Mastercard Foundation, warns that by 2030 women working in Africa’s outsourcing sector are more likely than men to see their tasks replaced by automation. It estimates that tasks performed by women are on average 10 percent more vulnerable to automation, and that up to 40 percent of human tasks in the sector could be automated. Lower-paid roles, which make up 68 percent of the workforce, are particularly at risk. Mastercard Foundation+1

In human terms, this means the very jobs pitched as ladders into the middle class — call-center agents, back-office clerks, junior analysts — may be the first to feel the impact of chatbots and automated workflows.

The gender angle is critical. If AI quietly automates the routine tasks concentrated in women’s jobs while preserving or upgrading mainly male roles, it will deepen inequalities that universities and corporate diversity programs have spent years trying to narrow.

2.2 Invisible labour in the AI machine

AI itself runs on human labour — but that labour is often invisible.

Major tech companies routinely outsource data labeling and content moderation to workers in Africa and Asia. A Brookings analysis notes that these workers, sometimes called the “ghost workers” of AI, often face low pay, limited benefits and long hours engaging with traumatic or disturbing content. Brookings+1

In Nairobi, content moderators working for a Meta subcontractor have sued over alleged human rights abuses, describing being paid as little as $1.50 an hour to watch videos of murders, rapes and other atrocities. The legal battle has raised sharp questions about whether Big Tech can be held accountable for conditions in outsourced facilities, and whether governments so hungry for digital jobs will defend workers or investors. TIME+1

As AI automates certain white-collar tasks in the North, it also creates a parallel labour market in the South — one that is poorly regulated, psychologically risky and easy to hide behind layers of contracting.

If the future of jobs in Africa is to be more than a story of digital sweatshops cleaning the world’s data for export, that will require legal, political and ethical choices, not just better software.

2.3 Banking, agriculture and other sectors in transition

Beyond outsourcing, AI is quietly reshaping work in sectors that dominate African economies:

• Banking and fintech: Nigerian and Kenyan banks are rolling out AI-powered chatbots, fraud detection and credit scoring. These tools can extend services to unbanked populations — but they also reduce demand for frontline staff while creating a small number of highly skilled data roles. optimusai.ai+1
• Agriculture: AI-driven advisory apps promise to help smallholder farmers optimize planting, access weather forecasts and link to markets. In practice, these tools may not replace farm labour but could change who has bargaining power, as those who control data and platforms capture a larger share of value. World Bank+1
• Manufacturing: Studies of South Africa’s apparel industry, for example, suggest that automation so far has had limited impact on employment, partly because the sector still relies heavily on manual skills and wages are low. ScienceDirect+1

In many cases, AI will not instantly destroy jobs. It will rearrange tasks inside jobs, alter wages, and shift risks onto workers who have little say in how technology is introduced.

3. Where new jobs may emerge

The story is not all loss.

A McKinsey analysis suggests that African economies could unlock up to $100 billion a year in value from generative AI alone across sectors like banking, retail, agriculture and the public sector — on top of gains from “traditional” machine learning. McKinsey & Company+1

New roles are already appearing:

• AI operations and maintenance: managing AI systems for banks, telecoms and governments; monitoring model performance; ensuring compliance with privacy and anti-bias rules.
• Local language and cultural adaptation: training and fine-tuning AI systems in Kiswahili, Yoruba, Amharic or isiZulu; curating local datasets; ensuring tools understand African contexts rather than defaulting to North American norms.
• Digital public infrastructure: building and running national ID systems, digital payment rails and e-government platforms that increasingly rely on AI for fraud detection, planning and service delivery.

The International Labour Organization, rethinking AI’s impact on the future of work, emphasizes that the biggest changes will likely involve how jobs are organized, paid and protected, not just how many jobs exist. It stresses that AI can create new tasks and roles even as it automates others — and that social dialogue and policy will shape the balance. International Labour Organization+1

For Africa, the question is whether these higher-value opportunities will be captured by a tiny elite of highly educated workers in capital cities, or whether they can be spread through broad-based upskilling and strategic public investment.

4. Supporters: AI as Africa’s leapfrog moment

Many African leaders and international organizations see AI as a chance not to fall further behind, but to leap ahead — or at least to catch up.

The World Bank report on the future of work in Africa argues that digital technologies, if combined with investments in human capital, could help countries move from low-productivity self-employment to more productive, formal jobs in services and manufacturing. World Bank+1

A Microsoft white paper on AI and the future of work in Africa highlights macroeconomic benefits, potential productivity gains, and workers’ own perspectives: many African employees see AI as a tool that can remove drudgery and open up more creative or complex work, provided they are trained and consulted in the process. Microsoft

The Mastercard Foundation, looking specifically at the outsourcing sector, insists that Africa can lead in responsible AI adoption, not just react to it, by investing in targeted upskilling programs — especially for women and youth — so that workers at risk of automation can “trade up” into better jobs. Mastercard Foundation+1

Supporters often make three key points:

1. Latecomer advantage: Because many African economies are not yet deeply locked into legacy systems, they can adopt AI in flexible ways, avoiding some of the organizational inertia that slows change in richer countries.
2. Demographic dividend: A young population, if given the right skills, could supply the global shortage of AI-savvy workers.
3. Problem-solving potential: AI could help address chronic shortages in health, education and public administration, amplifying the impact of limited human expertise.

From this angle, the real danger is not that AI will do too much in Africa, but that it will do too little — because infrastructure, skills and finance are missing.

5. Critics: Digital colonialism and the illusion of “neutral” progress

Skeptical voices in the Global South warn that AI, left to market forces, may simply repeat old patterns of extraction in new forms.

Researchers writing on AI and African development point out that most AI systems are designed, trained and owned by companies in the Global North, often using data collected in the South without meaningful consent or benefit sharing. They describe this as a new layer of “data colonialism,” where the raw material is not rubber or copper but human behaviour and language. arXiv+1

A Brookings essay, bluntly titled “AI is not Africa’s savior,” warns against technosolutionism — the belief that deploying AI can substitute for hard political and institutional reforms. It argues that without addressing weak infrastructure, limited social protection and governance deficits, AI could deepen inequality rather than reduce it. Brookings+1

Human-rights scholars examining AI in the Global South emphasize that automation often arrives in contexts already marked by labour exploitation, weak regulatory capacity and unequal access to justice. In such settings, AI can magnify existing power imbalances — for example, when algorithmic management in warehouses or call centers increases surveillance without improving wages or safety. IJHSSM+1

Critics do not argue that Africa should reject AI. They argue that who owns, governs and profits from AI matters as much as which tasks it automates. A future in which African workers are mostly data labelers, content moderators and low-level operators of foreign systems is not emancipation; it is a digital update of an old story.

6. Four choices that will shape the future of jobs

The future of jobs in the era of AI is not a natural disaster; it is a political and ethical project. For Africa and the broader Global South, four choices stand out.

6.1 From fear of automation to negotiation over its terms

The first choice is to move beyond a simple “jobs lost vs jobs gained” narrative and negotiate how automation happens.

That means:

• involving workers and unions in decisions about AI deployment;
• using labour law to ensure that productivity gains are shared through better wages, hours and protections;
• requiring impact assessments before large-scale automation in sectors like BPO, banking or logistics.

The International Labour Organization stresses that AI’s impact is as much about wages, conditions and bargaining power as about the raw number of jobs. International Labour Organization

Without institutions that can negotiate, African workers may experience AI as something done to them, not with them.

6.2 From digital sweatshops to decent digital work

The second choice is whether to accept AI-related jobs at any cost, or insist on minimum standards for digital labour.

Countries competing to attract content moderation or data-labeling contracts can easily slide into a race to the bottom on pay, mental health support and legal protections. The lawsuits in Kenya against Meta’s contractor, and the global reporting on traumatic working conditions in AI data pipelines, show the human cost of ignoring this. TIME+1

African governments and universities could instead collaborate on:

• ethical guidelines and certification for AI-related outsourcing,
• regional standards on pay, counselling and workplace transparency,
• public campaigns that make citizens aware of the hidden labour behind AI.

The goal should not be to close the door to digital work, but to raise its floor.

6.3 From imported platforms to local innovation

The third choice is about industrial policy: will African economies simply plug into foreign AI platforms, or also build their own?

Local innovation does not necessarily mean competing with Silicon Valley on foundational models. It can mean:

• building domain-specific AI for agriculture, health or education using African data and expertise;
• developing open-source tools and datasets in African languages;
• training a generation of AI engineers, designers and ethicists who are rooted in their societies.

The World Bank, McKinsey and others note that the macroeconomic value of AI in Africa could be large — but that realizing it depends on domestic capabilities, not just foreign investment. McKinsey & Company+2Microsoft+2

If Africa remains mainly a market and a data source for imported AI, the best jobs will stay elsewhere.

6.4 From emergency training to lifelong learning

Finally, AI will compress the half-life of skills. What a young person learns at 20 may be obsolete at 35.

This makes lifelong learning more than a slogan. It requires:

• rethinking universities and technical colleges as hubs for continuous upskilling,
• aligning curricula with both AI-related opportunities and the enduring need for human skills — care, negotiation, critical thinking, creativity,
• using AI itself as a tool to deliver flexible, low-cost learning in local languages, including for workers already in the informal sector.

Analysts of the future of work in emerging markets point out that countries with lower GDP per capita may face slower initial automation, but also have fewer resources to retrain workers once changes accelerate. RBC Gestion d’Actifs Globale+1

If Africa does not invest in lifelong learning now, the next wave of AI could lock millions into outdated skills.

7. A future still open — if Africa is allowed to write it

Seen from the Global North, the future of jobs in the AI era is often discussed in terms of preserving a middle class, reskilling white-collar professionals, or regulating powerful tech firms.

Seen from much of Africa, the baseline is different. It is about millions who have never had formal work, young people searching for a first job, women juggling unpaid care and precarious income, migrants moving between rural and urban economies, and workers in digital sweatshops whose fingerprints are all over the AI systems the world now celebrates.

AI will not automatically fix this. It could make it worse. But it also offers tools and leverage that previous generations did not have.

The question is whether African societies — workers, unions, universities, entrepreneurs, governments — will have the power and space to shape how AI is used, or whether those decisions will be made far away, in boardrooms and data centres where African lives are numbers on a dashboard.

The future of jobs in the era of AI is still open. For the Global South, and especially for Africa, the task is not just to survive that future, but to insist on being one of its authors.