AI turns disaster news into structured global risk maps
New approach turns worldwide disaster reporting into organised risk maps, highlighting connected consequences and improving insight into how crises evolve across countries and sectors.
A new AI-powered dataset developed by the European Commission’s Joint Research Centre, in cooperation with the Italian technology company Engineering Ingegneria Informatica and the Institute of Health and Society at the University of Louvain, is turning fragmented disaster reporting into structured knowledge to help researchers and policymakers better understand how crises unfold and interact.
The dataset covers more than 3,000 disaster events across 175 countries and 26 hazard types, drawing on global reporting to reduce geographical and thematic gaps in existing databases.
Climate and geological disasters, including floods, hurricanes, earthquakes, and wildfires, are processed into structured ‘storylines’ that trace causes, impacts, and responses.
A key feature of the system is its ability to identify cascading effects, in which one event triggers a chain of secondary impacts, such as infrastructure disruption, agricultural losses, or disease outbreaks.
Unlike traditional datasets that record impacts in isolation, the AI-generated knowledge graphs reveal interconnected risk dynamics that are often hidden in standard reporting.
The pipeline uses large language models and retrieval-augmented techniques to extract relevant articles and turn them into structured summaries and visual networks.
Why does it matter?
The development shifts disaster analysis away from fragmented reporting and towards more structured, interconnected intelligence. By showing how hazards cascade into broader social, economic, and environmental impacts, it can help policymakers and emergency services anticipate secondary risks more effectively, rather than reacting to isolated events.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!