Today’s UN report, “Leveraging AI to enhance multi-hazard early warning systems,” highlights technology’s crucial role in global safety. As extreme weather events intensify, Google actively supports the UN’s Early Warnings for All initiative. Our decade of AI breakthroughs in crisis resilience aims to prevent natural disasters from catching communities by surprise.
Google collaborates closely with the UN, governments, and international organizations to transform disaster preparedness. Our advanced AI tools are central to this global mission. We strive to make helpful information, from initial forecasts to real-time alerts, available to billions worldwide.
AI for Early Warnings: Prediction and Preparation
Timely forecasts empower proactive action before disaster strikes. During the 2025 hurricane season, Google’s WeatherNext model accurately predicted Hurricane Melissa’s Jamaican landfall five days in advance. This crucial lead time enabled the Jamaican Met Service to notify the public effectively.
In Nigeria, Google’s river flood forecasts drive impactful anticipatory action. UN OCHA launched a programme in Adamawa state, triggering early interventions like shelter preparation based on high-risk forecasts. GiveDirectly also used these forecasts in Kogi State, delivering cash transfers to help families evacuate and protect property.
Our Flood Hub now covers 2 billion people across over 150 countries at significant flood risk. We continuously improve forecasting with partners like the World Meteorological Organization (WMO). A multi-national pilot showed local data significantly enhances AI forecasts in ungauged areas, proving the synergy between global AI and local expertise.
To further research, Google open-sourced its Groundsource dataset for urban flash floods and its hydrology modeling framework. These resources help experts build new approaches while retaining local data control. The Czech Hydrometeorological Institute successfully integrated our framework, showcasing its global applicability.
For wildfires, Google uses satellite imagery to track fire boundaries in Search and Maps across 34 countries. In collaboration with the Earth Fire Alliance and Muon Space, we developed the FireSat constellation. Three new FireSat satellites recently launched, aiming for faster, unprecedented wildfire detection globally.
Real-Time Alerts and Post-Disaster Response
Access to reliable, authoritative information is critical during crises. In 2025, Google connected people with crisis information over 10 million times daily. Our Public Alerts system surfaces content from alerting authorities via Common Alerting Protocol (CAP) feeds.
These alerts, featuring data from over 90 countries, appear across Google Search, Maps, and Android notifications. This ensures vital public safety information, such as severe weather warnings, reaches affected individuals quickly. We encourage more nations to publish CAP alert feeds, expanding global coverage.
While earthquake warning remains challenging, Google has progressed in alerting those outside the epicenter. When devastating earthquakes struck Venezuela last month, the Android Earthquake alerting system proved invaluable. It alerted millions of users via Android phones acting as mini-seismometers, giving them precious seconds to take cover.
Post-disaster, rapidly delivering life-saving aid is paramount. AI-powered insights boost efficiency for governments and aid organizations. The Data Insights for Social and Humanitarian Action (DISHA) initiative, with Google and the UN Satellite Centre (UNOSAT), developed an advanced damage assessment workflow.
This workflow uses Open Buildings and Building Damage Assessment models to analyze satellite imagery, quickly identifying damaged structures. Deployed 11 times, it supports responses to earthquakes, floods, and cyclones. This system performs high-precision analysis of hundreds of thousands of buildings in short timeframes, saving UNOSAT specialists weeks.
For example, when Hurricane Melissa devastated Jamaica in October 2025, AI assigned preliminary damage scores to over 385,000 buildings, informing recovery efforts. After February 2026 floods in Colombia, UNOSAT rapidly assessed damaged infrastructure by cross-referencing AI-derived building maps with radar imagery. This informed humanitarian and national government response planning.
A Global Commitment for a Safer Future
Individual models offer powerful insights, but combining imagery, population, and environment data addresses complex real-world queries. Google’s Earth AI collection integrates our climate and geospatial models. This delivers actionable, planetary intelligence, supporting disaster response and comprehensive monitoring.
We are dedicated to advancing AI solutions and strengthening global partnerships. Together, we build a more resilient future, working towards our shared mission of protecting communities. Our journey continues, driven by innovation and a deep commitment to global crisis resilience.
Source: Google Blog (The Keyword)
