TLDR: China has launched MAZU-Urban, an AI-powered agent designed for urban multi-hazard early warning, at the 2025 World Artificial Intelligence Conference (WAIC). This system integrates advanced AI with meteorological science to enhance disaster prevention and mitigation efforts globally, with initial donations to Djibouti and Mongolia.
Shanghai, China – At the prestigious 2025 World Artificial Intelligence Conference (WAIC) on July 26, China officially unveiled MAZU-Urban, an innovative AI Agent for Multi-Hazard Early Warning. This groundbreaking system, developed jointly by the Shanghai Meteorological Service (SMS), the National Meteorological Centre of the China Meteorological Administration (CMA), China United Network Communications Co., Ltd. Shanghai Branch, the Shanghai Institute of Science and Intelligence, and the Shanghai Artificial Intelligence Laboratory, marks a significant leap in global disaster preparedness.
MAZU-Urban is designed to deeply integrate artificial intelligence technology with meteorological science, offering a comprehensive solution for urban multi-hazard early warnings. Dr. CHEN Zhenlin, Administrator of CMA, formally donated the MAZU-Urban system to representatives from Djibouti and Mongolia during the WAIC opening ceremony, witnessed by Prof. Celeste Saulo, Secretary-General of the World Meteorological Organization (WMO). This act underscores China’s commitment to global cooperation in meteorological disaster prevention and mitigation.
Feng Lei, director of the Shanghai Meteorological Service, highlighted the system’s core functionalities, which span four key areas. Firstly, it excels in disaster assessment and analysis, identifying disaster frequency and typical occurrence locations within a country. Secondly, MAZU-Urban enhances monitoring and early warning capabilities by leveraging China’s Fengyun meteorological satellites for global monitoring and integrating this data into numerical forecasting models, including the Fengqing large model developed by the CMA. This significantly improves forecasting capabilities for other nations.
The third key area focuses on the efficient issuance of warnings, ensuring that forecasts reach the general public promptly with clear instructions on necessary actions. Finally, the system incorporates a progressive response service mechanism, a model already implemented by the CMA and globally recognized, as part of its AI corpus material. This allows countries to rapidly establish national meteorological disaster response plans and standards based on AI-generated outputs, thereby accelerating institutional development.
MAZU-Urban’s smart architecture integrates host, tablet, and mobile devices into a ‘3-in-1’ framework. The all-in-one device serves meteorological and emergency management departments, providing minute-level disaster data collection and localized defense guidelines. The tablet device caters to industry-specific needs, such as port and shipping, by combining multi-source observations and user-defined thresholds. The mobile device targets the public, delivering location-based alerts, evacuation guidelines, emergency shelter navigation, and emergency contact information.
Also Read:
- China Advocates for New Global AI Governance Body Amidst Fragmented Landscape
- Exponential View Explores Superintelligence, AI Agents, and Future Global Challenges
Since January of this year, MAZU-Urban has undergone successful piloting in 35 countries and territories across Asia, Africa, and Oceania, receiving positive feedback. With the support of CMA, MAZU-Urban is the first globally shared product researched and developed by SMS. The Shanghai Meteorological Service plans to accelerate the joint construction of city-based meteorological early warning operational systems, contributing ‘Shanghai experience’ and ‘Chinese solutions’ to global meteorological disaster prevention and mitigation efforts.
