Expert Urges Govt On Metrics To Measure AI Strategy

As Nigeria strengthens its artificial intelligence policy framework, AI infrastructure expert Oluwaseyi Ayodeji has called for a more detailed national baseline to measure the country’s computing capacity, energy reliability, talent pool and progress in AI adoption.

Nigeria’s artificial intelligence ambitions must be backed by more precise measurements of its existing infrastructure, skills and capacity if the country is to effectively assess the implementation of its National AI Strategy, AI infrastructure expert, Oluwaseyi Ayodeji, has said.

Ayodeji, an AI infrastructure expert and founder of Regal Stack, a think tank focused on Africa’s sovereign participation in the global AI economy, said Nigeria’s policy architecture around artificial intelligence was beginning to take shape but needed a stronger baseline to determine what the country actually has and how much progress is being made.

His position comes as Nigeria continues to develop its AI ecosystem following the publication of the National AI Strategy in 2025 and the recent announcement of the National Digital Cloud Policy.

The National AI Strategy identifies foundational AI infrastructure, ecosystem development, sector adoption, responsible AI and governance among its key pillars, while the Digital Cloud Policy seeks to strengthen investment, cloud infrastructure, government adoption and digital sovereignty.

According to Ayodeji, however, having a strategy is only the starting point.

“The first version is almost never the final version,” he said, drawing from his experience of authoring strategy documents during his career.

He explained that strategic documents are often developed when information is incomplete, meaning that assumptions and broad targets may be necessary at the outset. However, he said, those assumptions should be tested and refined as more data becomes available.

“That principle matters enormously for AI,” Ayodeji said, stressing that Nigeria needs a baseline that establishes the country’s existing AI infrastructure rather than relying primarily on projections of what it hopes to build.

One area requiring greater precision, he said, is computing capacity.

While data-centre capacity is often discussed in terms of megawatts, Ayodeji noted that power capacity alone does not determine how much AI computing capability a country has.

“There are also important distinctions between hardware that has been procured, hardware that has been installed, hardware that is operational and hardware that researchers, startups and businesses can actually access,” he said.

He argued that Nigeria should be able to determine the amount of accelerator capacity installed and operational in the country, the location of such infrastructure, the classes of GPUs available and who can access them, including the cost and conditions attached to access.

Energy availability, he added, requires a similar level of scrutiny.

Using data from the Nigerian Electricity Regulatory Commission (NERC), Ayodeji pointed to the significant gap between installed and available electricity generation capacity. NERC’s April 2026 figures showed 13,625MW of installed grid-connected generation capacity, compared with average available capacity of about 4,286MW.

For AI infrastructure, he said, installed capacity does not necessarily translate into available or reliable capacity.

“Without that distinction, it becomes difficult to know whether the infrastructure being developed is actually sufficient for the ambitions of the strategy,” he said.

Ayodeji also called for greater granularity in the way Nigeria defines and measures AI talent.

He noted that programmes such as the Federal Government’s 3 Million Technical Talent (3MTT) initiative have expanded digital skills training across areas including artificial intelligence, cloud computing, data science, software development and cybersecurity.

However, he cautioned against treating everyone trained across these fields as a single category of “AI talent”.

Drawing on his experience in enterprise learning architecture and adult learning, Ayodeji said the skills required to support AI adoption vary considerably across industries and functions.

“A software engineer may need to learn how to use an AI coding assistant effectively. A finance professional may need to understand how AI can improve forecasting and analysis. A sales team may need a different set of capabilities altogether,” he said.

The same principle, he argued, should apply at the national level. Skills development should begin with the work the economy needs people to perform and then work backwards to identify the capabilities and training required.

Beyond conventional AI specialists such as machine-learning engineers and researchers, he said Nigeria would also need professionals capable of building and operating the physical infrastructure supporting the technology.

These include data-centre technicians, electrical and mechanical engineers, fibre specialists, controls engineers, commissioning professionals and construction workers with relevant technical capabilities.

Nigeria’s skills strategy, he argued, should therefore measure not only the number of people trained but also the specific roles required by the economy, existing shortages, geographic distribution and whether trained individuals are securing productive employment.

Ayodeji believes the National AI Strategy could become more effective by introducing a more comprehensive measurement framework, potentially through a national AI dashboard.

Such a system could track metrics including computing capacity, GPU availability, data-centre infrastructure, dependable power, connectivity, AI infrastructure investment, workforce categories, startup access and AI adoption across key sectors.

The objective, he stressed, should to establish a shared and reliable picture of the country’s progress.

“Every serious strategy needs a North Star,” Ayodeji said, adding that precise key performance indicators would enable policymakers and other stakeholders to identify what is working, where resources are being underutilised and where additional investment is required.

He argued that clearer targets would also make it easier for Nigeria to prioritise investments in infrastructure, skills and policy.

With AI infrastructure, technology costs, global supply chains, energy systems and workforce requirements evolving rapidly, Ayodeji said Nigeria’s AI strategy must also be capable of adapting.

“The goal is not simply to have an AI strategy. It is to build the measurement system that tells us whether the strategy is working,” he said.

Through Regal Stack, the think tank he founded, Ayodeji focuses on Africa’s participation and sovereignty in the global AI economy.

Nigeria’s next phase of AI development, he argues, should therefore move beyond policy ambition towards measurable implementation, establishing what exists, identifying what is missing and using reliable data to determine where investment and policy attention can deliver the greatest impact.