Nigeria Data Science Leader Urges AI Investment

By Salami Adeyinka

As Nigeria intensifies efforts to position itself as a regional technology hub, a leading data science professional, Itohowo Charles has called for investment in AI to compete globally.

He warned that the country’s artificial intelligence ambitions may falter without urgent investment in foundational data infrastructure.

The UK-based specialist argued that, while government agencies and technology firms announce AI initiatives with increasing frequency, the underlying data architecture required to support these systems remains fragmented and underdeveloped.

“Nigeria has brilliant technologists and a rapidly growing digital economy, but we’re building AI on shaky foundations,” Itohowo said in a recent technical briefing.

“Data fragmentation, poor interoperability, and weak governance frameworks mean that even well-funded AI projects struggle to deliver consistent results at scale.”

Itohowo’s intervention comes as the National Information Technology Development Agency (NITDA) prepares to implement elements of its National Artificial Intelligence Strategy, which identifies AI as critical to economic diversification and public service delivery.

However, he said that Nigeria’s current data environment presents significant obstacles. Key challenges he listed were data siloed across ministries and agencies with limited sharing protocols, inconsistent quality standards that compromise machine learning model training, absence of national interoperability frameworks, and gaps in privacy regulation and enforcement mechanisms.

“You can’t train reliable AI models on fragmented, inconsistent data,” he noted. “The technical reality is that machine learning systems are only as good as the data pipelines feeding them.”

Itohowo also pointed to countries that have treated data infrastructure as strategic national priority. Estonia’s X-Road platform, which enables secure data exchange across government and private sector systems, supports everything from healthcare to tax administration.

“Singapore’s National Digital Identity system provides a secure foundation for both public services and commercial innovation. Rwanda’s recent investments in digital ID and interoperability standards have enabled faster deployment of digital financial services.

“These aren’t just technology projects: they’re national infrastructure investments comparable to roads or electricity,” he explained. “Nigeria needs a similar commitment: a national data architecture, common interoperability standards, and governance structures that enable innovation while protecting citizens.”

The key lesson, he added, is sequencing. Countries that succeed invest in the boring but essential work first: data quality, standards, governance. The exciting AI applications come later, but they’re sustainable because the foundation is solid.”
Banking Fraud Detection: A Case Study Charles illustrated the practical consequences of infrastructure gaps using Nigeria’s banking sector, where fraud prevention remains a persistent challenge.

Legacy fraud detection systems deployed across Nigerian banks often generate false positive rates of 10 to 15 per cent, according to industry estimates. This means that for every 100 transactions flagged as suspicious, 10 to 15 are legitimate customer activities, creating operational overhead for banks and frustration for customers.

“Properly designed machine learning models trained on Nigerian transaction patterns could reduce fraud detection time from days to milliseconds,” Itohowo said. “In comparable deployments globally, false positives have been reduced by up to 60 percent, but this requires high-quality training data and well-engineered systems.”

He said the technical opportunity lies in data engineering pipelines that integrate multiple sources: payment rails operated by the Nigeria Inter-Bank Settlement System (NIBSS), credit bureau data, and internal bank transaction histories. When these streams are combined into what Charles describes as a “single source of truth,” fraud analytics can shift from a cost center to a value driver, enabling banks to approve more legitimate transactions faster while catching sophisticated fraud attempts.

However, achieving this requires interoperability standards that currently don’t exist at scale in Nigeria. “Without common data formats and secure sharing protocols, each bank builds isolated systems that can’t benefit from network-wide learning,” he noted.

The science data specialist outlined specific policy and investment priorities for government and private sector stakeholders such as the National Data Architecture to establish interoperability standards across government agencies and sectors, create secure data-sharing protocols that balance innovation with privacy protection, build national digital identity infrastructure linked to core services.

On quality and governance, he suggested developing data quality benchmarks and certification processes, implementing privacy-by-design principles in all new systems, and establishing clear data governance frameworks with accountability mechanisms.

Others he outlined included skills and capacity, to invest in technical training for data engineering and AI governance roles, supporting public-private partnerships that build institutional capability, and creating incentives for data professionals to contribute to national projects.

Measurement and accountability. “The investments needed aren’t primarily about AI algorithms,” Itohowo emphasised. “They’re about data plumbing, standards development, and governance structures: less glamorous but far more impactful.”

“Nigeria doesn’t lack AI ambition,” he acknowledged. “What we need now is the discipline to build the infrastructure first, measure results honestly, and hold ourselves accountable for delivery. The countries winning the AI race understood this years ago. We can still catch up, but only if we’re honest about where the real work needs to happen.”