Science Talk: Innovations in AI for Agri-Food Systems Research for Development

  • Date
  • Time
    10:30 am > 03:30 pm UTC+03:00
  • Location
    United Nations, Conference Room 3, Nairobi, Kenya

Thematic Areas: Climate Action, Better Crops, Capacity Sharing, Digital & Data

This event will feature three sessions that showcase CGIARs research for development activities that apply AI for accelerating agrifood transformation and impacts. The first session will delve into different AI applications, showcasing advancements in crop diversity conservation, genebank utilization, crop breeding, and monitoring of crop productivity and climate adaptation. The second session will focus on generative AI applications, highlighting innovative solutions including AI-powered chatbots and solutions for capacity sharing to support extension services and small-scale producers. The final session will focus on ethical AI, featuring interactive discussions on ensuring responsible AI development and deployment in agri-food systems. 

Session 1 (10:30-12:30): AI Applications for Accelerating Agri-Food Systems Research 


  • AI Applications for Crop Diversity Conservation and Genebank Use (Charlotte Lusty, CGIAR Senior Director – Genebanks; Venuprasad Ramaiah, IRRI): Genebanks are a key resource to conserve, manage and access crop diversity for coping with climate change, yet they are largely underutilized mainly due to lack of information on usefulness of the accessions. By application of AI-based methodologies CGIAR fast-tracks the utilization of genebanks, reducing cost and time to 1/16th and 1/10th, respectively, of the ‘business as usual’ approach. This presentation will increase awareness of this approach. 
  • Leading AI in Agriculture: Advancements in Crop Breeding and Protection (Michael Selvaraj, Alliance Bioversity-CIAT): This session will highlight the latest findings and applications of AI in crop breeding, showcasing how these innovations contribute to accelerating development of improved varieties and gaining efficiencies along the way.  AI-Enabled Remote Sensing for Measuring Productivity, Emissions and Adaptation (Charles Spillane, University of Galway): The Tracking Adaptation in Agricultural Systems (TAPAS) project ( combines remote sensing and AI for measurement of agricultural systems. TAPAS is an AI for Societal Good research project that arose as a collaboration between Uni of Galway and Alliance Bioversity CIAT. The TAPAS project has developed AI-enabled satellite remote sensing methods for monitoring agri-food systems, with a particular focus on enabling measurement of emissions and measurement of climate change adaptation, spatially and temporally. The TAPAS project is working in partnership with a range of CGIAR Centers and Initiatives to co-develop use cases where AI-enabled satellite remote sensing can measure effectiveness of agriculture productivity, mitigation and adaptation interventions. 

Panel discussion 

    • How Can AI Accelerate the Agri-Food Systems Transformation in a Climate Crisis? (All Session 1 Speakers): This panel discussion will explore the cutting-edge developments in the innovative use of AI within CGIAR research, focusing on its potential to transform agri-food systems in the ongoing climate crisis. Panelists will discuss the latest advancements in AI applications that aim to enhance agricultural productivity, climate resilience, and sustainability. They will also delve into the critical challenges that need to be addressed to scale these AI solutions effectively.

Session 2 (13:20-14:50): Enhancing Capacity Sharing with Generative AI Applications 


  • The WhatsApp-Based AI-Powered Agri-Food Systems Academy (Andrea Gardeazabal, CIMMYT): The need for capacity building in farming communities is significant. ICT enables distance learning, but low digital literacy is a barrier in South Asia. Leveraging WhatsApp’s widespread use, this workshop will demonstrate AI-powered learning tools, train participants on AgroTutor Academy’s Learning Management System, and co-design micro-courses on agri-food systems. 
  • DINA: The AI-powered evidence clearing house on digital agri-food systems (Satish Nagaraji, CIMMYT): Adopting digital innovations in agri-food systems faces multi-dimensional challenges. Solid evidence on technology and social aspects aids adoption. The CGIAR’s Digital Innovation Navigation Assistant (DINA), an AI tool that leverages multiple Large Language Model (LLMs), provides this evidence. This session will discuss the need for such evidence, share experiences with generative AI, and showcase results via a hands-on demo. 
  • Supporting Extension Services Using AI Chatbots with CGIAR Knowledge (Jawoo Koo, IFPRI): Language Model-based AI chatbots have demonstrated potential in supporting agricultural extension services by summarizing complex scientific literature and presenting it in a locally relevant context. However, concerns remain about the accuracy of AI-generated content. This presentation will showcase recent progress by CGIAR and technical partners to ensure AI chatbot responses are accurately grounded in scientific knowledge.

Panel discussion 

  • Can Generative AI Help Small-Scale Producers in the Global South Thrive? (All Session 2 Speakers): This panel will explore how generative AI can support small-scale producers in the Global South by localizing AI applications for capacity sharing with CGIAR partners. Panelists will discuss interrelated challenges in scientific evaluation, infrastructure provision, data governance, and digital literacy. They will also highlight opportunities for cooperation in syndicating content, establishing standardized benchmarking for performance evaluation, and creating mechanisms for user feedback to continually improve AI system performance.

Session 3 (14:50-15:30): Ethical AI in Agri-Food Systems 


  • Design Thinking for Ethical AI in Agri-Food Systems (Andrea Gardeazabal, CIMMYT) using Challenges from AI-griculture Hackathon (Alessandra Furtado, CIP): The interactive session will engage participants in identifying ethical challenges and generating solutions for implementing AI in agrifood systems. The methodology follows a condensed design thinking approach with four phases: Empathize, Define, Ideate and Share. Participants will work in small groups to ideate, discuss ethical implications and share solutions, i. The session aims to encourage creative thinking and collaboration to raise awareness of ethical considerations in the design and deployment of AI technologies for agri-food systems. 


  • Jawoo Koo (IFPRI) 
  • Andrea Gardeazabal (CIMMYT) 
  • Charlotte Lusty (CGIAR Senior Director – Genebanks) 
  • Venuprasad Ramaiah (IRRI) 
  • Charles Spillane (University of Galway) 
  • Michael Geever (University of Galway) 
  • Satish Nagaraji (CIMMYT) 
  • Michael Selvaraj (Alliance Bioversity-CIAT) 
  • Alessandra Furtado (CIP)