The integration of artificial intelligence (AI) technologies into agriculture holds urgent and transformative potential for enhancing food security across Sub-Saharan Africa (SSA), a region acutely impacted by climate change and resource constraints. This paper examines experiences from the Artificial Intelligence for Agriculture and Food Systems (AI4AFS) Innovation Research Network, which provided funding to innovative projects in eight SSA countries. Through a set of case studies, we explore AI-driven solutions for pest and disease detection across crops such as cashew, maize, tomato, and cassava, including a real-time health monitoring tool for Nsukka Yellow pepper. Using participatory design, and key informant interview, robust monitoring and evaluation, and incorporating ethical frameworks, the research prioritizes gender equality, social inclusion, and environmental sustainability in AI development and deployment. Our results demonstrate that responsible AI practices can significantly enhance agricultural productivity while maintaining low carbon footprints. This research offers a unique, localized perspective on AI’s role in addressing SSA’s agricultural challenges, with implications for global food security as demand rises and environmental resources shrink. Key recommendations include establishing robust policy frameworks, strengthening capacity-building efforts, and securing sustainable funding mechanisms to support long-term AI adoption. This work provides the global community, policymakers, and stakeholders with critical insights on establishing ethical, responsible, and inclusive AI practices that can be adapted to similar agricultural contexts worldwide, contributing to sustainable food systems on an international scale.

Bioethanol production from cellulosic materials is important in mitigating the concomitant displacement and exploitation of primary food crops for biofuel production and reducing carbon emissions which exacerbate climate change. The problem of reduced yield in the production and availability of yeast locally poses a barrier to market adoption and penetration of bioethanol. The study examined the effect of particle size and different yeast strains on the yield of bioethanol from waste sawdust that was sourced from a local timber processing centre. The samples of yeast were prepared from bakers yeast (Saccharomyces cerevisiae) and palm wine yeast (Saccharomyces chevalieri). The sawdust was reduced to 212μm, 300μm, and 500μm particle sizes. The samples of each particle size were pretreated and hydrolyzed with H₂SO₄ and fermented with S. cerevisiae or S. chevalieri yeast. The results obtained show that the weight, pH, density, viscosity, flash point, and heating value of the produced bioethanol ranged between 221.67 and 322.64g, 6.2 and 6.6, 0.821 and 0.878g/mL, 1.073 and 1.193, 14 and 16C, and 20.5 and 23.1MJ/kg, respectively, while the alcohol content of each of the samples was 69. Furthermore, the bioethanol yield from Saccharomyces cerevisiae yeast was 213.9mL, 193.2mL, and 186.3mL, for the 212μm, 300μm, and 500μm particles, while for Saccharomyces chevalieri yeast, the yield was 289.8mL, 255.3mL, and 220.8mL for the 212μm, 300μm, and 500μm, respectively. An ANOVA on the effect of particle size on ethanol yield shows a significant difference at 5 level of significance. The study demonstrated that the use of locally produced yeast and increasing the surface area of sawdust increase bioethanol yield. Hence, it was concluded that better yeast strain use and biomass particle size reduction to a level that allows the optimal surface area for the reaction improve the yield of bioethanol. The study outcome can help in ameliorating the continued dependence on fossil fuels and the food security problems arising from displacing or utilizing food for fuel and could also encourage commercial-scale cellulosic ethanol production from waste.