context: At the 2026 World AI Conference, the PRC announced plans to expand the AI-powered weather early warning solution Mazu to 30 countries over the next five years. By combining satellite observation, artificial intelligence models and climate data, the system aims to improve weather risk management in agriculture and support more resilient food production.
Extreme weather has become a growing challenge for agricultural production worldwide. Rising temperatures, more frequent droughts, floods and sudden cold spells are increasing uncertainty for farmers and affecting crop yields. For agriculture, where production decisions are closely linked to weather conditions, earlier and more accurate forecasting has become an increasingly important tool for reducing losses and improving efficiency.
At the 2026 WAIC (World AI Conference), the PRC introduced Mazu, an AI-based weather early warning solution designed to connect monitoring, forecasting, early warning and agricultural services. The system integrates data from Fengyun weather satellites, AI forecasting models and digital platforms to provide customised solutions for different sectors, including agriculture, transport, energy and the low-altitude economy.
Unlike traditional weather forecasting systems that mainly provide general information, Mazu focuses on application-based services. By combining weather information with local production conditions, the system can provide more targeted support for specific crops and regions. According to available information, Mazu has already supported applications in more than 40 countries, with global crop yield forecasting accuracy exceeding 94 percent and coverage across 15 major agricultural countries on six continents.
Agriculture is one of the key areas where the technology is expected to create value. In Uzbekistan, where apple production is an important agricultural sector, Chinese meteorological teams have used Mazu to establish orchard weather service demonstration sites and develop frost warning solutions. These services help farmers prepare for sudden temperature drops and reduce potential crop losses. In Tajikistan, where drought poses a major challenge to wheat production, the system provides customised weather information to support local food security.
The development of Mazu also reflects a wider shift in agricultural technology. As climate risks become more difficult to predict, agriculture is moving from relying mainly on experience-based decisions towards data-driven management. AI-powered forecasting can help farmers optimise planting schedules, improve disaster preparation and strengthen production stability.
By making advanced weather information more accessible, Mazu represents a new approach to applying AI in agriculture. The expansion announced at WAIC shows how AI technologies are increasingly being used to address real-world challenges, from climate adaptation to global food security.