Shanghai launches the country's first fire safety corpus foundation and open-source plan

Shanghai released the nation’s first opensource fire safety corpus base plan at the 2026 World Artificial Intelligence Conference, and simultaneously rolled out a number of phased achievements in AI + fire safety. The event was held in Shanghai on July 20 and was jointly announced by the Shanghai Fire and Rescue Bureau together with multiple organizations. According to the introduction, this fire safety corpus base covers business areas such as fire safety theory, laws and regulations, fire prevention supervision, and fire suppression operations. It brings together a large amount of corpus data, Q&A data, chainofthought data, and evaluation data, and is supported by a knowledge graph and a professional knowledge base, aiming to provide foundational support for model training and applications in the fire safety field. At the same time, Shanghai also released the achievements of building an alldomain, allvolume data resource pool, the “Silu” fire safety AI standard system, and a cluster of fire safety intelligent agents, and launched a call for submissions for AI + fire safety intelligent scenarios, promoting application implementation in areas such as early warning and prediction, supervision and law enforcement, public fire safety management, education and publicity, and fire suppression and rescue. On the same day, the new unmanned fire equipment application research and testing center was also officially inaugurated, exploring the use of technologies such as humanoid robots in specialized fire safety scenarios.