Alibaba Damo Academy unveils AI model that can reportedly forecast major weather events up to 12 months ahead

Announcement at COP30

Alibaba Damo Academy (阿里巴巴达摩院) used its invited slot at the 30th session of the United Nations Climate Change Conference (COP30) to showcase AI tools designed for climate adaptation and disaster response. The centerpiece was its Baguan ("八观") meteorological large model, which the academy said has been upgraded to provide long-range climate warnings — reportedly able to predict major meteorological events such as El Niño and severe cold waves up to 12 months in advance. A bold claim. A potential game‑changer for planning and relief operations?

Performance and deployments

It has been reported that the Baguan model has already been fielded in Zhejiang, Shandong and Beijing, with deployments focused on near-term extreme weather forecasting. According to Alibaba, the system reduced typhoon intensity forecast errors by more than 50% in those localized applications. The model mixes short-term storm forecasting with longer-horizon climate signals to give authorities more lead time for evacuation, infrastructure protection and economic planning — though independent verification of the 12‑month horizon has not been published.

Geopolitical context and implications

The announcement comes as China’s tech firms push software and algorithmic advances amid U.S.-led export controls on advanced semiconductors and AI chips, a shift that has increased emphasis on domestic model development and cloud services. At COP30, the pitch was clear: better forecasts mean fewer lives lost and less economic disruption. Will international agencies and other countries adopt or validate China-developed models? For now, Alibaba’s claims add to a wider debate about how commercial AI tools can be trusted, shared and governed in a geopolitically fractured data and chip landscape.