AI-Enabled Climate-Smart Agriculture for Smallholder Resilience in Sub -Saharan Africa: A Systematic Review
Adeyemi Olatunbosun
*
Robinson College of Business, Georgia State University, Atlanta Georgia, USA.
Emmanuel Oluwasegun ISMAILA
Project Center for Agro Technologies, Skolkovo Institute of Science and Technology, Moscow, Russian Federation.
Mariam Iyabo Adeoba
Department of Mechanical, Bioresources and Biomedical Engineering, University of South Africa, South Africa.
Olaitan Ebenezer Oluwadare
Applied Mathematics Department, School of Physics, Engineering, Mathematics and Computer Science, Delaware State University, USA.
Tope Joseph Arayombo
Department of Agricultural Leadership, Education and Communication, University of Georgia, USA.
Adedamola Oladunni
Department of Nuclear, Plasma & Radiological Engineering, University of Illinois Urbana-Champaign, USA.
Oluwafunmilayo Esther Ajiferuke
Department of Agricultural Leadership, Education and Communication, University of Georgia, USA.
Manfred Obinwanne Igwenagu
Natural Resources and Environmental Sciences, Prairie View A & M University, USA.
*Author to whom correspondence should be addressed.
Abstract
Background: Smallholder farmers in Sub-Saharan Africa are highly vulnerable to climate change impacts, including droughts, floods, and pest outbreaks, which threaten yields, incomes, and food security. Climate-smart agriculture (CSA) aims to enhance productivity, adaptation/resilience, and mitigation, yet adoption remains limited due to resource constraints. Artificial intelligence (AI) tools come as a game-changer to make CSA more accessible and effective for farmers with small farms.
Aim: This systematic review synthesises empirical evidence on AI-enabled CSA interventions and their impacts on smallholder resilience in Sub-Saharan Africa.
Methods: PRISMA 2020 guidelines were followed, with searches on Scopus, Web of Science, CAB Abstracts, Google Scholar, and AGRICOLA for studies published between 2015 and 2025. Eligible records were those with evidence of AI applications linked to CSA practices and resilience outcomes in smallholders. After screening 1,301 records, nine studies/reports made it to the final inclusion and were assessed using the Mixed Methods Appraisal Tool (MMAT). Results were synthesised narratively.
Results: The nine included sources (covering Kenya, Tanzania, Mali, Ethiopia, Malawi, Nigeria, and Ghana) describe four main AI intervention types: image-based pest/disease detection, advisory chatbots, precision resource management apps, and predictive modelling. These models have already been used by hundreds to tens of thousands of farmers. Reported impacts include yield increases (e.g., +0.9–1 t/ha for rice, up to 25% for wheat), income gains ($200–600/ha), reduced input waste, and improved adaptation to shocks. All interventions align strongly with CSA pillars, particularly productivity and adaptation, with mitigation benefits from efficiency gains.
Conclusion: AI-enabled CSA shows that it can be transformative for enhancing smallholder resilience in Sub-Saharan Africa, delivering modest but meaningful short-term gains in yields, incomes, and climate adaptation. Effects are positive, although it depends on the context of application, constrained by digital barriers. Policymakers and developers should prioritise inclusive, mobile-based scaling. Further long-term, rigorous studies are essential to confirm durability and broaden geographic coverage.
Keywords: Artificial intelligence, climate-smart agriculture, smallholder farmers, Sub-Saharan Africa, digital agriculture, advisory systems