Talented Nigerians, outdated opportunities

On September 24, in a room on the sidelines of the United Nations General Assembly, Nigeria told the world it was the next frontier for global outsourcing.

The Hire From Nigeria campaign, launched in Abuja four weeks earlier and taken to New York by the office of the Vice President, set an initial target of one million export-linked jobs by 2030.

The trade minister, Jumoke Oduwole, was careful to say this was a market-facing economic proposition rather than a communications campaign.

Buried in the same material was the number that should have led every report. Nigeria currently has an estimated 20,000 people working for domestic and international outsourcing firms.

One million from 20,000 in four years is a 50-fold increase. It is worth pausing to consider how that compares with the rest of the continent, because Nigeria is not entering an empty market.

Egypt employed 195,300 people in exporting outsourcing firms as of June 2026, with exports of 5.2 billion dollars in the last financial year, according to the chief executive of its IT industry agency.

Morocco had 148,500 offshoring jobs in 2024 and is targeting 270,000 by 2030. Ghana, with a population a 15th of Nigeria’s, has about the same 20,000 jobs Nigeria has and has set itself a target of 100,000.

So Nigeria, starting from Ghana’s base, has announced a target four times Morocco’s and five times Egypt’s current workforce, in half the time either took to build theirs.

I want this to work. I am Nigerian, I have delivered technology programmes in Nigeria at national scale, and I now spend my life training people into exactly the kind of work this campaign is chasing.

But the numbers in front of us describe a strategy that is entering the market at the wrong rung of the ladder, at the precise moment that rung is being sawn off.

Start with the most instructive comparison, which is not Egypt but South Africa. South Africa runs the best-organised business services industry on the continent.

In 2025, it created 26,346 jobs serving international markets, its strongest year since 2018. Roughly nine in 10 of those jobs went to people under 35. By any measure, it is succeeding.

And yet the head of its industry body, BPESA, now expects the sector to add about 150,000 jobs by 2030, against the 500,000 envisaged in the national master plan agreed in 2021. The best-run outsourcing industry in Africa, in a growth year, is revising its own target down by roughly two-thirds.

That is the benchmark against which Nigeria’s one million should be read. Not as cynicism about Nigerian capability, but as evidence about what this industry delivers when it is working well.

The most quoted number in this debate is that artificial intelligence will eliminate 40 per cent of Africa’s outsourcing jobs. That is not what the research says, and the misquotation matters.

The study, produced by Caribou and Genesis Analytics with the Mastercard Foundation and published in April 2025, is a task-level analysis. It finds that close to 40 per cent of tasks in the sector could be automated by 2030. Tasks are not jobs. A role can lose half its tasks and survive, changed.

But read further, and the picture sharpens in a way that is specifically uncomfortable for Nigeria’s plan. Entry-level roles make up 68 per cent of the sector’s workforce, and more than half of entry-level tasks are automatable.

Customer experience work, which accounts for 44 per cent of employment, has about half its tasks exposed. And tasks performed by women are, on average, 10 per cent more susceptible to automation than those performed by men, because women are concentrated in the junior roles.

So the risk is not evenly spread across the industry. It is concentrated in the entry level, in voice work, and in the roles women hold. Which is to say, precisely the segment a country recruiting a million new workers from a standing start would have to fill first.

The honest counter-argument deserves space here because it is strong. Egypt has more than doubled its outsourcing workforce since 2021, in full view of the same technology. South Africa had a record year in 2025.

The same Caribou research found that most workers in the sector see artificial intelligence as an enabler rather than a threat. Automation has not emptied African contact centres, and there is no evidence it is about to.

The clearest job losses so far came from a different direction entirely. In April 2026, Sama issued redundancy notices to 1,108 staff in Nairobi after Meta ended a content and data annotation workstream.

Those jobs did not vanish because a model replaced them. They vanished because one client changed its mind. That is concentration risk, and it is the characteristic danger of building an economy on being somebody else’s back office.

The detail that should most trouble Nigerian policymakers is not Egypt’s size. It is Egypt’s direction.

Ahmed El-Zaher, who runs ITIDA, argues that artificial intelligence will transform the nature of outsourcing work rather than simply eliminate it, and he is explicit that Egypt’s ambition is to become a regional base that develops and exports AI solutions rather than one that supplies services.

Egypt is not defending the bottom of the value chain. It is climbing out of it while its headcount still grows.

It can afford to climb because of what sits underneath. More than 800,000 university graduates enter its system annually, and the state is targeting 800,000 ICT-trained workers in 2026 alone.

At the November 2025 Global Offshoring Summit, ITIDA signed memoranda with 55 global firms, expected to create 70,000 to 75,000 jobs over three years. Concentrix alone intends to grow its Egyptian headcount to around 35,000 by the end of 2028.

One multinational, in one country, plans more outsourcing jobs than Nigeria’s entire sector currently holds.

Nigeria’s deeper problem is not ambition. It is order of operations. The National Artificial Intelligence Strategy was published in September 2025, built around five pillars and 34 strategies.

Related News

• Reactions as Tinubu Chinese tablets go viral
• Debt worse than leprosy, Obasanjo replies Tinubu
• AI images: Atiku demands prosecution of INEC officials

Independent analysis of it found that none of those 34 strategies carries a binding key performance indicator, and that as of March 2026, the strategy’s own capstone institution, an independent AI Governance Regulatory Body, had been proposed but not constituted.

The accompanying code of practice had not been finalised. Nigeria spends about 0.2 per cent of GDP on research and development, against a global average above two per cent.

Compare Ghana. It launched its National AI Strategy in April 2026 with a commitment to establish a Responsible AI Authority within a year, cabinet approval for 250 million dollars for a national AI compute centre, and a jobs target of 100,000. Strategy, then regulator, then funded compute, then a target proportionate to the base.

Nigeria has run the sequence backwards. It has launched the recruitment campaign before the regulator exists, before the national fibre spine is laid, and before compute at meaningful scale is available.

This is the pattern I see repeatedly in technology programmes, and it is why so many of them disappoint. The failure is almost never in the technology. It is in the institutional layer around it: the decisions nobody has made, the body nobody has constituted, the question of who is accountable when something goes wrong. Strategy documents are cheap. Regulators, substations and fibre are not.

Nigeria’s compute position is difficult to state precisely, which is itself revealing. Published counts of operational data centres range from 11 to 25, depending on the source and the definition, and capacity estimates run from roughly 30 megawatts to 86.

For continental context, the African Data Centres Association put the whole of Africa at 360 megawatts of active capacity in early 2026, which is about 0.6 per cent of the global total.

There has been real movement this year. Kasi Cloud commissioned what it describes as West Africa’s first hyperscale-ready, AI-capable campus in Lagos in May 2026, backed by the sovereign wealth fund, with its own 132-kilovolt substation and a path to 100 megawatts.

MTN opened a 150-million-dollar facility in Ikeja. Airtel’s Eko Atlantic data centre has taken delivery of its first GPUs. In August, MTN confirmed a partnership with a UAE-backed platform targeting 150 megawatts of AI-ready capacity across Nigeria and South Africa.

But the binding constraint is not buildings; it is electricity. Nigeria has over 13,000 megawatts of installed generation capacity and was averaging around 4,200 megawatts available in April 2026. That is why every credible facility builds its own substation or captive gas supply, and why data centre promoters are being courted by the gas industry rather than the grid.

Kenya provides the cautionary version. The one-billion-dollar Microsoft and G42 geothermal data centre at Olkaria, announced in 2024, has been paused.

President Ruto explained that switching on that single facility would require cutting power to half the country. Announced capacity and deliverable capacity are different things across this continent, and the gap is measured in electrons rather than press releases.

Connectivity shows the same split. Nigeria has deployed over 101,000 kilometres of fibre, but fewer than 265,000 fibre-to-the-home subscriptions, with nearly a fifth of the backbone concentrated in Lagos and Abuja, and more than 155,000 fibre cuts recorded in April and May 2026 alone.

Project BRIDGE, which is meant to add 90,000 kilometres of open-access fibre to every local government area, was due to begin in late 2025 and, according to the governing party’s chairman, was starting physical deployment this month.

The ownership pattern now forming is worth naming plainly because it will determine where the value lands.

Cassava Technologies is building Africa’s AI factory capacity in partnership with Nvidia, in a deal worth up to 720 million dollars, with thousands of GPUs deployed in South Africa and expansion planned into Nigeria, Kenya, Egypt and Morocco.

In October, the World Bank announced an AI infrastructure initiative with Cassava and Nvidia as knowledge partners, offering governments shared, pay-as-you-go access to compute. Gulf capital is behind the paused Kenyan project and the MTN platform. Nigerian sovereign money sits inside Kasi Cloud.

So the owners of African compute are emerging as telcos, pan-African infrastructure groups, sovereign funds and Gulf investors. The chips are American. The base models are American too: N-ATLAS, Nigeria’s own multilingual model covering Yoruba, Hausa, Igbo and Nigerian-accented English, is a genuine achievement, and it is fine-tuned from Meta’s Llama 3.

That is not a criticism. Fine-tuning on a foreign base model is the rational move for a country without a national GPU cluster. But it should end the pretence that Nigeria is competing at the frontier. Nigeria’s realistic position is in adaptation, deployment and integration, and those are respectable places to be. They are also where the margin sits in every mature technology market.

India built the model Nigeria is now trying to join. It is instructive to look at what that model is doing in 2026. Infosys reported revenue of 20.16 billion dollars for its last financial year, growth of 3.1 per cent in constant currency, and in July cut its guidance for the current year to between 1.5 and 3 per cent.

Artificial intelligence accounted for 8.2 per cent of revenue in the first quarter. Management has pointed to AI productivity effects and cautious client spending.

India’s services giants are repricing themselves away from billable headcount towards platform outcomes. A strategy built on counting Nigerian heads is entering the business model India is currently exiting.

Four things, none of which require abandoning the ambition.

Aim higher up the stack. Point Hire From Nigeria at AI operations, model evaluation and quality assurance, data engineering, regulated back-office work and Nigerian-language AI services built on N-ATLAS. Avoid competing for voice-heavy customer experience work, where half of all tasks are automatable and where Egypt, Morocco and South Africa already have a decade of scale on us.

Constitute the regulator. A strategy without its regulator is a document. Every month the governance body remains unconstituted is a month in which the institutions that already hold Nigeria’s most valuable data, the banks, the telcos and the national identity system, set the terms by default.

Measure placement, not training. The flagship 3MTT programme has trained somewhere between 135,000 and 160,000 fellows, depending on which official source you read, against a target of three million by 2027, with several thousand placed in jobs. The constraint is not the supply of trained Nigerians. It is demand. The Mastercard Foundation, a partner in this very campaign, says as much: train to demand rather than training first and looking for jobs afterwards.

Publish the baseline. Nobody can tell you how the figure of 20,000 was arrived at. If the country is going to be held to a million by 2030, it deserves a counted starting point and an annual published figure.

Nigeria does not have a talent problem. It has more young people with more appetite for this work than any country on the continent, and I would back them against any comparable workforce anywhere. What it has is a timing problem and an institutional one, and those are both fixable, but only by someone willing to say out loud that the plan as currently pitched sells the wrong decade of work.

Talent is everywhere. Opportunity is not. The difference between them, this time, is whether we build the ladder before we invite a million people to climb it.

Areola is a certified senior project manager and founder of Skills Co-op.