Leveraging artificial intelligence to strengthen nutrition program monitoring and evaluation in Nigeria

¹College of Business Administration, Colorado State University, Fort Collins, Colorado, United States, ²Department of Biology, University of Arkansas at Little Rock, Arkansas, United States, ³Faculty of Environmental Design and Management, Department of Urban and Regional Planning, Obafemi Awolowo University, Ile-Ife, Nigeria, ⁴Department of Civil Engineering, University of Ilorin, Ilorin, Nigeria, ⁵Department of Public Health, University of Dundee, Scotland, United Kingdom, ⁶Bristol Medical School, University of Bristol, Bristol, England

^&Corresponding author
Yetunde Olaogun, College of Business Administration, Colorado State University, Fort Collins, Colorado, United States

Commentary    

Nigeria faces one of the world's most severe burdens of child malnutrition [1,2]. United Nations Children's Fund (UNICEF) estimates that the country has the second-highest number of stunted children globally, with malnutrition implicated in about 45% of deaths among children under five [1]. The burden is heaviest in the conflict-affected north, where stunting exceeds 60% in some states, but no region is spared. Successive governments, United Nations agencies, and non-governmental partners have responded with a broad portfolio of nutrition interventions, the largest of which, the National Home-Grown School Feeding Programme (NHGSFP), has grown from about one million to nearly 10 million children since 2016 [3].

The value of these investments depends on how well they are monitored and evaluated. Monitoring and evaluation (M&E) is the cornerstone of evidence-informed decision-making [4]. It shows whether programmes achieve their intended outcomes, surfaces implementation problems, and guides continuous improvement [4]. In Nigeria, however, M&E remains largely manual, relying on paper-based reporting, periodic supervisory visits, and delayed data flows that often reveal problems only after the moment for corrective action has passed. The result is a system that struggles to generate the real-time evidence that large-scale nutrition programmes need.

Artificial intelligence (AI) offers a way to close this gap. Across government functions, AI and advanced analytics are increasingly used to process data automatically, analyse beneficiary feedback, and generate real-time performance insights, not to replace human judgement but to sharpen it [5]. International agencies are beginning to formalize this shift; UNICEF's evaluation office, for example, has issued guidance on using machine learning to strengthen evaluation practice [6].

Nutrition programmes are especially well suited to these tools because they must track both long-term outcomes and fast-moving processes such as food availability, supply chains, funding flows, and service delivery. In Ethiopia, UNICEF's end-user monitoring system has shown how real-time digital data can evaluate nutrition service coverage and quality across more than 500 districts, providing timely evidence to improve delivery [7]. Building on such systems, AI can flag emerging problems, analyse large volumes of qualitative feedback from beneficiaries and frontline workers, and predict where operational failures are most likely, so that limited supervisory resources are directed where they matter most [8] For Nigeria, where nutrition programmes run nationally but still depend on manual monitoring, AI could automate routine data processing, identify bottlenecks in near real time, and produce actionable insight to strengthen accountability.

Nigeria has taken an important first step. In 2025, it launched a National Artificial Intelligence Strategy that names social services, agriculture, and healthcare as priority sectors [9], providing a policy foundation for applying AI to M&E. Translating that vision into routine practice, however, faces real obstacles. Many nutrition programmes still rely on paper-based or fragmented digital systems that limit the high-quality data AI requires; digital infrastructure, including reliable connectivity and electricity, is uneven, particularly in the rural communities where malnutrition is highest. Technical capacity to build and interpret AI-enabled M&E is scarce, and governance frameworks for data quality, privacy, and ethical use are still evolving. Nigeria's experience with earlier national digital initiatives shows that ambitious strategies do not translate automatically into effective implementation.

Closing the gap between policy and practice will therefore require more than a strategy. First, government should establish clear governance structures, timelines, and accountability mechanisms to operationalize the strategy, with sustained political commitment across federal, state, and local levels, since nutrition programmes are delivered largely subnationally. Second, because AI is only as good as the data behind it, Nigeria should invest in reliable electricity, broadband, interoperable information systems, and standardized data collection, progressively replacing paper-based reporting with digital platforms. Third, applied training should equip M&E professionals, programme managers, and policymakers to use AI for data management, predictive analytics, and decision support. Fourth, local research and innovation ecosystems, including university-based AI centres and implementation research embedded within government programmes, should be strengthened so that solutions are locally owned. Finally, sustained financing and partnerships among government, universities, development partners, and civil society will be needed to move beyond isolated pilots.

Throughout, ethics must remain central. As AI enters nutrition M&E, safeguards should protect privacy, promote transparency, and limit algorithmic bias. Nigeria's Data Protection Act (2023) offers a legal foundation, but AI-specific guidance for public-health and nutrition programmes, covering transparent algorithms, routine bias assessment, and human oversight of AI-supported decisions, is still needed [10]. Crucially, AI should complement, not replace, foundational investment in routine data systems; its greatest value comes where it is layered onto digitized monitoring rather than deployed over fragmented paper records. Understood this way, AI is not a shortcut but a stage in building resilient M&E. If Nigeria embeds it within stronger systems and keeps human judgement, equity, and accountability at the centre, it can make its nutrition programmes more responsive, improve child-health outcomes, and accelerate progress toward the Sustainable Development Goals on hunger and healthy lives.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

All the authors have read and approved the final version of this manuscript.

References