Nigeria’s AI future at risk without indigenous data, expert warns

Folalumi Alaran in Abuja

A Tech expert, Oladayo Olasupo, has warned that Nigeria’s ambition to lead in artificial intelligence (AI) could be at risk due to poor data infrastructure and dependence on foreign-trained models, unless urgent action is taken to build indigenous data systems.

Oladayo said while government efforts such as the National AI Strategy and recent funding for startups reflect commendable efforts, they will have limited impact unless backed by high-quality, locally sourced data to train machine learning (ML) systems.

According to him, in a statement on Tuesday, many AI tools currently adopted in Nigeria are built on datasets that reflect foreign realities.

This, he noted, results in “tools that do not understand our languages, miss our context, and even reinforce bias against local users.”

The expert noted that the risks go beyond technical limitations and include ethical, cultural, and economic consequences.

He pointed to sectors such as healthcare, agriculture, and education, where Nigeria stands to benefit the most from AI but only if the systems are trained on accurate local datasets.

He said: “In Nigeria, a country with a population of over 200 million and a vibrant cultural and linguistic diversity, locally relevant data is essential for building AI systems that resonate with its unique socio-economic and cultural contexts.

“From healthcare to agriculture, education to fintech, AI applications tailored to Nigeria’s needs such as climate tech for flood prediction or vernacular AI for its 500+ languages require datasets that reflect local realities.

“LLMs trained on non-local data may fail to understand Nigerian languages, idioms, or cultural nuances. For instance, a chatbot trained on Western datasets might struggle to process requests in pidgin English or Hausa, limiting its utility.

“Non-local datasets can embed biases that perpetuate stereotypes or discriminate against Nigerian users. For example, an AI hiring tool trained on global data might favor candidates with Western qualifications, marginalizing local talent.

Oladayo referenced recent remarks by the Minister of Communications, Innovation and Digital Economy, Dr. Bosun Tijani, who noted that Nigeria is among the top 60 countries globally in AI readiness and is developing a homegrown large language model (LLM), a commendable step.

While commending the effort, the tech expert said such milestones will remain symbolic unless backed by strategic investment in local data infrastructure. “

“However, without high-quality, localized data, these systems risk being irrelevant or ineffective. Data-driven AI can transform sectors by enabling predictive analytics, automating processes, and fostering innovation, but its efficacy hinges on the quality, relevance, and volume of data used for training”, he said

He further noted that over 500 languages spoken across Nigeria present both a challenge and an opportunity for vernacular AI, local voice applications, and culturally attuned services.

He added: “High-quality, clean, and structured datasets are scarce in Nigeria. Many sectors, including healthcare and education, lack digitized, comprehensive data, limiting the ability to train effective AI models. The UN Economic Commission for Africa notes that African data ecosystems are in nascent stages, with private sector involvement still emerging. Poor data quality often incomplete, inconsistent, or biased leads to the garbage in, garbage out problem, where flawed inputs produce unreliable outputs.

“Also, Nigeria’s digital infrastructure, including internet access and data storage, remains inadequate in many regions. Rural areas, in particular, face connectivity challenges, hindering data collection efforts. Secure and reliable data storage systems are also limited, increasing risks of data breaches or loss.

“The absence of robust data governance frameworks leads to siloed, inconsistent, and error-prone datasets. Without standardized protocols for data collection, cleaning, and validation, organizations struggle to prepare ML-ready data. Data poisoning, where malicious or misleading information is introduced into datasets, is another concern, potentially distorting AI outcomes.

Olasupo urged the government to treat data as national infrastructure and to invest heavily in digitization, cloud storage, and AI-ready data pipelines.

He also called for new regulatory frameworks focused on AI transparency, bias, and privacy.

“Nigeria stands at a pivotal moment in its AI journey. By addressing data bottlenecks and prioritizing localized datasets, the country can unlock the full potential of AI to tackle its unique challenges.

“Nigeria’s young, tech-savvy population is both a source of AI talent and a market for its applications, with the right data, we can build solutions that truly transform lives.

“The government, private sector, and academia must continue to collaborate to build a data-driven AI ecosystem that is inclusive, ethical, and impactful. By increasing investment in infrastructure, skills, and governance, Nigeria cannot only overcome the challenges of clean, ML-ready data but also set a global standard for AI that reflects its rich diversity and boundless potential”, he added.