Everyone is right, yet everything is wrong.

On the same week in May 2026, two stories broke in Nigerian tech circles and landed with predictable force.

The first: Kled AI, a US-based data marketplace that pays users to upload photos and videos for AI training, announced it had removed its app from Nigeria's App Store and implemented a full IP ban on the country. The reason — approximately 95% of activity from Nigerian users was fraudulent. Black screens. Duplicate files. AI-generated images submitted as original content. And, in the incident that apparently broke the founder's patience, thousands of fake Japanese passports with Nigerian faces photoshopped onto them, flooding the platform's KYC system over a single weekend. The company had paid out hundreds of thousands of dollars over four months before the scale of abuse became financially unsustainable. For comparison, Malaysia, Indonesia, and the Philippines — with ten times the user base — had fraud rates below 10%.

The second: Tosin Eniolorunda, CEO of Moniepoint, Nigeria's foremost fintech unicorn, disclosed at a public event that his company currently has approximately 500 open vacancies it cannot fill. Not for lack of applicants — for lack of candidates who meet global standards. Moniepoint had committed in 2024 to hiring exclusively within Nigeria. By 2025, that commitment was straining under the weight of a talent supply that keeps leaving. "Not only could we not find people at the quality and the quantity we needed," Eniolorunda said, "the few people that we found were not up to the global standards that we need."

Both stories generated the familiar social media response: outrage, denial, a split between people admitting the problem and people accusing the messenger of bad faith. Both conversations collapsed quickly into the usual arguments about character, culture, and whether Nigerians are being unfairly maligned.

I'm not particularly interested in that argument. What I find interesting is why both things are true simultaneously — and what the system producing them actually looks like.

Breeding Cobras

In colonial India, the British government grew concerned about the population of venomous cobras in Delhi, so they came up with an elegant solution: offer a cash bounty for every dead cobra brought in.

The program worked initially, till it didn't. Entrepreneurial locals, recognizing a reliable income stream, began breeding cobras specifically to claim the bounty. When the government eventually discovered the scheme and cancelled the program, the breeders released their now-worthless stock, and the cobra population ended up larger than before the intervention.

This is what happens when incentive structures are deployed without the institutional context that makes honest participation the rational choice. When the gap between the reward and the legitimate path to it is wide enough, and when there is no functioning mechanism to make exploitation costly, exploitation becomes the logical outcome. This is not a moral observation. It is a systems observation.

Kled arrived in Nigeria with an incentive structure — upload data, earn money — and no scaffolding. Not because Kled was negligent, but because the scaffolding it assumed would be there, the digital identity infrastructure, functioning fraud enforcement, a population for whom gaming a micro-gig platform represents a worse option than legitimate participation, simply does not exist in any reliable form. The platform met the conditions it met, and produced the output those conditions predictably generate.

The Moniepoint story is the same mechanism running in reverse. The skill exists, and Nigerian engineers are, by global market evidence, genuinely competitive — US, UK, European, and Canadian companies are actively building core product teams with Nigerian talent, not outsourcing peripheral tasks but embedding Nigerian developers at the center of their product work.

The cobra effect here is that the naira has lost approximately half its value against the dollar since early 2024. For any engineer on a fixed naira salary, that is a 50% pay cut in real terms, delivered silently, without negotiation, and so the rational response to that environment is not to stay, on the contrary… staying is the irrational choice.

Two populations.

Two rational responses to the same broken incentive structure. One games foreign platforms because no legitimate pathway exists. The other leaves for foreign markets because staying is economically self-defeating.

Both are exits that silently compound the problem.

The Cancellation

Here is what happens when you run both exits simultaneously in the same ecosystem.

The skilled population leaves, and with it goes the mentorship pipeline, the senior engineering culture, the institutional knowledge that compounds inside organizations over time. Moniepoint's 500 vacancies are a measurement of how much of that pipeline has already walked out. The companies that remain compete for a shrinking pool of mid-level talent, driving up costs for smaller startups and leaving a growing tier of roles simply unfilled.

The population without those skills, facing no legitimate digital pathway and a collapsing cost of living, engages with whatever foreign income opportunity presents itself. Some of that engagement is legitimate. Much of it, as Kled's numbers suggest, is not — because when the most accessible digital income mechanism is a poorly-guarded foreign platform and the cost of exploitation is near-zero, exploitation becomes the path of least resistance.

The fraud generates IP bans. The IP bans raise the baseline suspicion level for Nigerian users across a growing class of global digital services. The next platform arrives in Nigeria pre-loaded with skepticism, tighter fraud thresholds, and less tolerance for anomalies, and the window for legitimate participation narrows even further.

The skilled population's exit means there is no local talent pipeline that could have produced legitimate high-quality contributors for platforms like Kled. The unskilled population's fraud means the platforms that could have generated income for both populations close the door faster.

Scattered, desperate efforts canceling each other out due to the compounding logic of a broken system with no architecture underneath it.

Eniolorunda diagnosed this as cultural decline. "The level that people are reasoning in this country is not as high as it used to be," he said, attributing the talent shortage to weaknesses in the education system, social media distraction, and what he described as a growing preference for quick money over career development.

It is worth noting that traditional Nigerian banks recruit year-round without lamenting the talent supply. The difference is not that banks attract a different kind of Nigerian. It is that banks pay at rates that make staying locally competitive with leaving.

The vacancy is in the architecture, not the Nigerian intellect.

The Ministry of Exits

Nigeria has a Federal Ministry of Communications, Innovation and Digital Economy. Its mandate, as stated, is to develop the country's digital infrastructure, grow the digital economy, and build Nigeria's capacity to compete globally in the technology sector. Its flagship initiative, launched with considerable fanfare, was the 3MTT programme — Three Million Technical Talent — a plan to train three million Nigerians to be globally competitive in tech.

On the FMCIDE's official YouTube channel, under the 3MTT banner, in a formal partnership with a dollar-account fintech, the Ministry hosted a webinar. The title: "Remote Work: The fastest path to earning in Dollars."

This is not a third-party event loosely affiliated with the Ministry.

This is the Ministry's own program, on its own channel, pointing its trainees toward the individual exit — earn in foreign currency, work for foreign employers, extract value from foreign economies — as the explicit goal of the training.

The institution mandated to build Nigeria's digital economy was running a program oriented toward helping Nigerians leave it.

The practical consequence of this, in retrospect, is almost too neat. 3MTT trains people toward remote dollar-earning. The trained talent takes remote jobs with foreign employers. Moniepoint, competing for the same talent on naira salaries in a collapsed currency, cannot fill 500 roles. The population 3MTT didn't reach, or didn't place, encounters Kled and games it. Nigeria gets IP-banned. The Ministry announces the next initiative.

What the Ministry could have done — should have done — is less visible and less filmable. Compute infrastructure is not a press conference. Subsidized or free access to the processing power that AI development requires does not generate a headline. A facility where developers can experiment with building systems rather than just using them does not trend on X. But it is the difference between a country that participates in the AI economy as a builder and one that participates as a labour reservoir for other people's products.

The electricity problem is real and belongs to a different ministry. But compute does not require solving electricity first. It requires solving prioritisation.

Instead, the Ministry operated at the announcement layer — the press releases, the webinars, the partnership launches, the global forums attended and addressed. The announcement layer functions perfectly. It generates coverage, signals intent, and produces the visible markers of activity.

What it does not produce is infrastructure. And infrastructure, specifically its absence, is what connects the Kled story to the Moniepoint story to the Ministry's record in a single unbroken line.

The Arithmetic

Reputation damage from incidents like Kled's exit is real, but it is the least important consequence. Reputation can, in theory, be rebuilt. The more durable cost is positional.

The global AI economy is not a future event. It is being constructed now. The decisions about which countries are inside the data layer, which have the engineering capacity to compete at the model layer, which have the infrastructure to build at the application layer — these decisions are being made in this window, through a combination of investment flows, talent distribution, platform access, and institutional capacity.

They accumulate slowly, quietly, completely skipping the empty announcements made for short term gratification.

A country that arrives at this moment without retained senior engineering talent, without compute infrastructure, without AI development ecosystems, and with a worsening trust profile in global digital markets does not simply miss a cycle. It starts off the next cycle from a worse starting position — with more reputational debt to service, a wider skills gap to close, and fewer tools available for either. The compounding runs in both directions.

The Kled ban will fade from the news cycle within days. Moniepoint will eventually adjust its hiring strategy. The Ministry will announce something new. None of that addresses what is actually happening, which is not a reputational crisis or a talent shortage or a fraud problem.

We're a country running on borrowed position — of global goodwill, raw talent, and digital access — without reinvesting in any of it, and that arithmetic only runs one way.