Assessment of climate risks in agriculture using AI

The FAO report, covering the period from 1991 to 2023, presents data on global annual agricultural losses from natural disasters amounting to 99 billion dollars. In India, extreme weather conditions in 2025 damaged at least 17.41 million hectares of sown areas. This figure is approximately 4.3 times higher than the damage area for 2024 and nearly nine times higher than for 2022.

For farmers, floods, cyclones, hail, and heatwaves result in ruined fields, flooded storage facilities, a lost growing season, and debt obligations. At the same time, existing approaches to climate risk assessment are often trapped between cumbersome reports with delayed data and complex meteorological models that fail to provide clear solutions for farms.

The authors of the article Samhita (Co-founder & CEO) and Gokul (Senior Research Associate, Resilience AI) note that natural hazard science and available satellite data have existed for a long time. To improve risk management, it is proposed to use specialized 24/7 access platforms. Such systems can apply machine learning and artificial intelligence to integrate data layers on hazards, land plots, infrastructure facilities, and loss history. This allows farming associations and cooperatives to receive up-to-date information within 72 hours and take practical measures to protect the harvest, livestock, and storage facilities before weather events escalate into natural disasters.