At present, Artificial Intelligence (AI) technologies in meteorology are developing rapidly, demonstrating significant potential in meteorological observation, weather and climate prediction, and weather service. However, a convergence of challenges including technological gaps, lagging regulation, data imbalance, limited generalization ability for extreme events, insufficient explainability, and inadequate public-private collaboration, poses severe challenges and potential risks to global development of meteorological AI. In this background, the World Meteorological Organization (WMO) convened the WMO AI Conference, themed "AI for Weather Prediction: Advances, Challenges & Future Outlook", from 9 to 11 September 2025 in Abu Dhabi, United Arab Emirates. The conference provided a comprehensive and in-depth discussion of the opportunities and challenges associated with the development of AI technologies in meteorology. A WMO AI Conference statement was released, which identified key priority areas for future global collaboration in meteorological AI. These areas include strengthening data foundations and infrastructure to ensure the availability of data for AI; establishing trustworthy evaluation and explainability frameworks; advancing capacity building and fit-for-purpose governance frameworks; fostering cross-sector collaboration among public, private, and academia; and developing common principles and guidelines for the ethical application of AI in meteorological services. The conference laid the groundwork for future global development and international cooperation in the application of AI within the meteorological community.

This article reviews advances in monitoring and nowcasting of severe convective weather (SCW), along with developments in operational nowcasting systems. It focuses on deep learning (DL)-based techniques using multisource data, highlighting associated challenges and opportunities. Based on multisource observations including those from dual-polarization weather radars and geostationary satellites, the monitoring capabilities of SCW types and intensities, convective initiation, and identification and tracking of convective storm cells have been significantly improved using advanced technologies, including storm structural feature recognition, fuzzy logic, and DL. Among these approaches, deep generative models have proven particularly effective, substantially improving the accuracy and extending the lead time of SCW nowcasting. The performance of the China Meteorological Administration's Severe Weather Analysis and Forecasting (SWAN) 3.0 system continues to advance, with widespread operational adoption across China. Future efforts will leverage higher-resolution observations and numerical weather prediction products at the hundred-meter resolution to enhance the understanding of the underlying mechanisms of SCW development at meso-γ- and microscales. Current purely data-driven AI models are transitioning toward physics-informed frameworks for SCW nowcasting. Integrating forecasters' operational expertise with state-of-the-art AI technology will further enhance operational capabilities in monitoring and nowcasting extreme SCW events.

The global AI (Artificial Intelligence) governance is currently evolving rapidly as the international community are jointly exploring the construction of a comprehensive and interconnected governance ecosystem. The United Nations has played a pioneering role in consolidating the core position of global AI governance, issuing a series of universal ethics and rules aimed at establishing a consensus framework. Leading AI technology entities such as the United States, the European Union, and the United Kingdom are accelerating the strategic layout of AI governance, striving to shape a coordinated and highly collaborative international governance system. Meanwhile, they are also competing for international discourse power in the field of AI governance. China has taken solid steps on the path of AI governance. By proposing the "Global AI Governance Initiative", China not only demonstrates a comprehensive perspective of development, security, and governance theoretically, but also promotes the transition of AI governance from rule-making to practical implementation through specific measures such as algorithm registration, assessment and evaluation, and post-event traceability checks. International organizations' AI governance norms and ethical initiatives for specialized fields such as education, medical care, and health also provide strong regulation and support for the development of related industries. Despite the above efforts, the governance of AI in the meteorological field on a global scale is still in a nascent state. Domestic and international efforts to establish guidelines, methods, and regulations for AI governance in the meteorological field are just beginning, and regulations and system standards are urgently needed. This article reviews the current status of AI governance in the United Nations and major AI technology countries and related international organizations, analyzes the unique challenges of AI application risks and governance in the meteorological field, and provides a global perspective for constructing an AI governance system in the meteorological community.

This article reviews advances in monitoring and nowcasting of severe convective weather (SCW), along with developments in operational nowcasting systems. It focuses on deep learning (DL)-based techniques using multisource data, highlighting associated challenges and opportunities. Based on multisource observations including those from dual-polarization weather radars and geostationary satellites, the monitoring capabilities of SCW types and intensities, convective initiation, and identification and tracking of convective storm cells have been significantly improved using advanced technologies, including storm structural feature recognition, fuzzy logic, and DL. Among these approaches, deep generative models have proven particularly effective, substantially improving the accuracy and extending the lead time of SCW nowcasting. The performance of the China Meteorological Administration’s Severe Weather Analysis and Forecasting (SWAN) 3.0 system continues to advance, with widespread operational adoption across China. Future efforts will leverage higher-resolution observations and numerical weather prediction products at the hundred-meter resolution to enhance the understanding of the underlying mechanisms of SCW development at meso-γ- and microscales. Current purely data-driven AI models are transitioning toward physics-informed frameworks for SCW nowcasting. Integrating forecasters’ operational expertise with state-of-the-art AI technology will further enhance operational capabilities in monitoring and nowcasting extreme SCW events.