China will deploy its AI powered meteorological early warning system, MAZU, in 30 developing countries over the next five years, strengthening global disaster preparedness and supporting the United Nations' Early Warnings for All initiative.

Developed by the China Meteorological Administration (CMA), MAZU combines AI forecasting models, Fengyun meteorological satellite data and local weather observations to deliver more accurate forecasts and tailored early warning services. Named after the ancient Chinese sea guardian Mazu, the platform supports more than 200 applications across sectors including agriculture, transport, energy and aviation.

The system is already operating in countries including Pakistan, Ethiopia, Sri Lanka and Djibouti, while meteorological agencies in more than 40 other countries and regions access it through cloud services.

An upgraded version, unveiled at the 2026 World Artificial Intelligence Conference in Shanghai, improves forecast resolution from nine kilometres to three kilometres and enhances extreme weather monitoring using AI, phased array radar and satellite technology. China has also trained nearly 1,000 meteorological professionals from more than 100 developing countries and regions to strengthen global early warning capabilities.