Report by Debola Ibiyode, founder and executive director of the AI (Adebola Ibiyode) Empowerment Foundation, and founder and CEO of CarbonAI | A review of artificial intelligence adoption across Nigeria paints a picture of a nation racing to use AI, while still struggling to build it (The study was conducted by OLGNova and AI in Action Now an initiative by AI Empowerment Foundation)
Scroll through your phone, walk into a hospital, visit a government office, or ask a Nigerian smallholder farmer about their planting decisions. Chances are, artificial intelligence is already part of the picture. Nigeria’s engagement with AI has exploded in recent years, and by most measures, the country is one of Africa’s most active adopters of the technology.
But a new macro and micro economic review of AI adoption in Nigeria raises a question that the headline numbers do not answer: is Nigeria truly adopting AI, or is it simply using it?
The difference, it turns out, matters enormously.
The Scale of Nigeria’s AI Engagement is Real
The numbers are striking. More than 70% of Nigerians have interacted with generative AI tools in some form (1). A survey of Nigerian organisations found that 93% have begun some form of AI adoption, with nearly a third reporting advanced integration (2). Nigeria’s AI research output has grown to over 11,676 scholarly works, with the University of Ibadan leading institutional contributions (3).
Across sectors, the footprint is visible:
Governance: The Federal Civil Service launched Service Wise GPT in January 2025, an AI tool for policy drafting, regulatory interpretation, and administrative task automation (4). Lagos and Abuja are piloting AI powered traffic management systems that have reportedly cut peak hour travel times by around 20% (5).
Healthcare: AI is helping address Nigeria’s severe health worker shortage. EepochZero trains AI models to detect radiological pathologies (6). Ubenwa, a locally developed solution, uses infant cry patterns to diagnose birth asphyxia (7). AI chatbots like FriendnPal and Chat Kemi are providing mental health support in a country with only 262 psychiatrists for over 240 million people (8).
Agriculture: The Nigeria Agricultural Intelligence Platform connects farmers, researchers, and policymakers through AI enabled advisory services (9). Studies show AI tools can increase crop yields by 20 to 30%, reduce crop losses by 26 to 31%, and cut input costs by up to 30% for farmers with access (10).
SMEs and Industry: AI enabled demand forecasting is cutting food waste in small businesses by 35 to 40%, and AI based route optimisation is reducing delivery times by up to 30% (11). Major manufacturers including Dangote Cement and Nigerian Breweries are using AI for predictive maintenance and production optimisation (12).
The policy environment is also moving. Nigeria has developed a National Artificial Intelligence Strategy (13), signed the Bletchley Declaration on AI safety (14), enacted a Data Protection Act (15), and established the National Centre for Artificial Intelligence and Robotics (13).
The Hidden Problem: Nigeria Uses AI, But Does Not Own It
Here is where the review delivers its most important finding. Despite widespread usage, most AI systems operating in Nigeria are developed and hosted outside the country.
The review frames this as the difference between AI consumption and AI creation. Consumption means using tools built elsewhere, such as the frontier AI models, cloud platforms, and diagnostic algorithms overwhelmingly developed in the United States, China, and Europe. Creation means building the models, datasets, and computational infrastructure domestically.
By this measure, Nigeria sits firmly at the consumption end of the spectrum. And this has consequences.
When a country consumes AI without creating it, value flows outward. Data generated by Nigerian farmers, patients, and businesses feeds systems whose intellectual property belongs to foreign entities. When the servers go down or the licensing terms change, local users have little recourse. The productivity gains are real, but the long term competitiveness gains are not.
The review finds that a 1% increase in AI adoption is linked to approximately a 0.65% rise in long run economic growth in Nigeria (16). That is a meaningful figure. But it also notes that AI could affect 4 to 5% of GDP related tasks by the end of the decade, with productivity improvements of around 10 to 15% in directly impacted areas (16).
Who Benefits and Who Does Not
Even the consumption gains are not equally distributed. The review documents a clear pattern: the benefits of AI adoption in Nigeria are concentrated among those who are already better connected.
Farmers with smartphones and reliable internet access are using AI advisory platforms to increase yields and reduce losses. The majority of Nigeria’s smallholder farmers, who form the backbone of the agricultural economy, remain outside the AI ecosystem entirely. They are constrained by lack of devices, poor connectivity, and limited digital literacy. Only 29% of commercial and agritech linked farmers actively use AI tools (10).
In healthcare, 90.8% of healthcare students believe AI will improve efficiency, but only 43.4% have had any practical exposure to AI tools (17). In education, 82.4% of students in one survey reported improved academic performance after using AI tools, but only 34.4% had reliable access to the infrastructure needed to use them (18).
At the enterprise level, large corporations are integrating AI rapidly. Small and medium enterprises, which form the vast majority of Nigeria’s productive economy, are falling behind, constrained by costs, skill gaps, and weak data infrastructure (12).
The labour market picture is similarly mixed. Estimates suggest that by 2030, around 9 million jobs could be displaced by automation, particularly in banking, public administration, and routine task sectors, while approximately 11 million new AI enabled jobs may emerge (19). Accessing those new jobs, however, requires skills that most of the current workforce does not yet have.
A Language Problem That Often Goes Unmentioned
The review also highlights a dimension of AI inequality that rarely makes the headlines:
language.
The majority of large language models are trained predominantly on English language data. This means AI tools perform worse, sometimes significantly worse, in Hausa, Yoruba, Igbo, and Nigeria’s hundreds of other languages (20). In healthcare, where accurate communication is critical, this is not a minor technical inconvenience. It is a structural barrier to equitable access.
One of the review’s clearer recommendations is targeted investment in Nigerian language datasets and locally grounded AI models. An initiative called Awarri is already working to build foundational AI datasets reflecting African linguistic and cultural realities (21). This kind of work needs to scale, and it needs funding.
What Nigeria Needs to Do Next
The review does not argue that Nigeria should slow its AI adoption. The productivity gains, even in consumption mode, are real and meaningful. But it argues compellingly that adoption without domestic capability building is a path toward permanent technological dependency.
Its recommendations centre on a transition from consumption to creation.
At the national level: Move from strategy to execution. Invest in sovereign AI infrastructure including domestically trained models, national computing capacity, and open source systems tailored to priority sectors like health, agriculture, and education. Align AI policy with industrial policy and labour market planning so that productivity gains translate into domestic value creation (13).
At the institutional level: Universities, hospitals, and government agencies should not just use AI. They should be building it. Research universities should establish innovation hubs. Healthcare and agricultural AI systems should be designed with rural and low income communities in mind from the outset, not as an afterthought.
At the community level: Expand AI literacy programs in rural and underserved communities. Support small and medium enterprises through adoption grants, low interest financing, and digital clusters. Treat AI skills as a labour market priority, with reskilling programs aligned to the needs of the digital economy (19).
On language: Localise AI systems. Integrate Nigerian languages, cultural context, and indigenous knowledge into model development. Fund language datasets and annotation infrastructure (20).
The Bottom Line
Nigeria is not failing at AI. It is doing something more specific and more complicated: it is adopting AI faster than it is building the institutional, infrastructural, and human foundations needed to sustain and own that adoption.
The review describes Nigeria’s AI trajectory as being in transition rather than maturity. That is an honest assessment. The country is engaging seriously with a transformative technology at a moment when the window to shape that technology’s development remains open. Whether Nigeria will move from being a consumer of AI to a creator of it depends on decisions being made right now about investment, policy, education, and inclusion.
The tools are there. The data is being generated. The talent exists. What is needed is the strategic intent to turn engagement into ownership, and the institutional commitment to follow through.
This article is based on a peer reviewed narrative review: “Adoption of Artificial Intelligence in Nigeria: A Macro and Micro Economic Review.” The study synthesised evidence from 41 sources across academic literature, institutional reports, and policy documents.
The study was conducted by OLGNova and AI in Action Now