In September 2025, Nigeria released its National Artificial Intelligence Strategy, a document developed under the guidance of the National Centre for AI Research and designed to chart the country’s course through the AI era until 2030. It covers data governance, ethical frameworks, innovation incentives, regulatory approaches, and plans for building AI capacity across sectors from agriculture to financial services. It is thorough. It is well-formatted. It cites the right international frameworks. And it has a problem that the document itself does not acknowledge.
The companies most involved in shaping that strategy are the same companies whose business models depend on extracting value from Nigerian data, Nigerian labour, and Nigerian markets. They co-authored the rules that will govern them. The strategy that is supposed to protect Nigerian interests in the AI era was substantially influenced by entities with a financial interest in those protections being as weak as possible.
This is not unique to Nigeria. Researchers tracking AI policy development across Africa have documented the same pattern in multiple countries. The government announces an AI strategy, technology companies are invited to contribute to the development process, the technology companies contribute language that is structurally favourable to technology companies, and the resulting document is published as national policy. The policy then creates the regulatory environment in which those same companies operate. The cycle closes on itself, and the citizens the policy is supposed to serve are not meaningfully represented in the process that produced it.
The implementation record makes the problem worse. Between December 2024 and October 2025, multiple Nigerian states experienced cabinet reshuffles affecting the ministries responsible for technology and innovation. Ekiti State’s flagship Ekiti Knowledge Zone, a project central to regional digital economy ambitions, had already accumulated a twelve-year delay before the latest round of political disruption added to it. Twelve years. For a project that was supposed to be a proof point for Nigeria’s technology future.
Implementation in Nigeria is not primarily a technical problem. It is a political one. The country consistently produces competent professionals who can write credible strategy documents. It does not produce the political continuity, institutional stability, or accountability structures that allow strategy documents to become functioning programmes. The pattern repeats across administrations, across sectors, and across decades. The strategy gets written, the launch event happens, the press releases go out, and then progress stalls at the point where it would require difficult political choices or sustained institutional commitment.
A 2025 independent review of Nigeria’s AI development initiatives found that existing programmes had produced limited impact. That phrase requires unpacking. Limited impact means the infrastructure situation is largely unchanged. The skills gap is largely unchanged. The talent is largely still leaving. The regulatory environment is still uncertain enough to make serious long-term private investment difficult. The strategies that preceded this one did not produce what they promised, and there is no structural reason to expect the new one will be different without the underlying governance conditions changing.
What would an honest AI strategy for Nigeria look like? It would start by naming the binding constraints rather than assuming them away. Not ethical AI frameworks adapted from European documents. Not data governance principles that make sense for countries with functioning data infrastructure. The binding constraints in Nigeria in 2026 are electricity availability, education quality at the secondary and tertiary levels, institutional stability within the agencies responsible for implementation, and a legal and regulatory environment stable enough to attract the kind of patient capital that technology infrastructure requires.
A strategy that begins with those constraints would be shorter, less impressive-looking, and more honest. It would describe a realistic five-year programme of foundational work before the AI ambitions it currently announces as near-term goals become achievable. It would commit to specific measurable outcomes with consequences attached to missing them. It would disclose who shaped the document and what interests they represented.
It would also be a document that the companies who currently shape these processes would not want written, because their influence over the process depends on the process remaining opaque and the goals remaining ambitious enough that accountability for outcomes is permanently deferred.
Nigeria has been scheduling its digital future for 2030 since at least 2010. When 2030 arrives, the target will move to 2035. The strategy documents will grow longer. The implementation gap will remain. And the companies that wrote the rules will continue operating comfortably within them.