Ko Barret, Deputy Secretary-General, WMO
Kamal Kishore, Assistant Secretary-General, UNDRR
Xavier Castellanos Mosquera, Under Secretary General, IFRC
Representatives of other partners,
Distinguished delegates, ladies and gentlemen,
It is an honor to address you today on a mission that lies at the very heart of the International Telecommunication Union’s commitment to saving lives: the use of Artificial Intelligence to power early warning systems for everyone, everywhere.
In 2022, the UN Secretary-General launched the Early Warnings for All (EW4All) initiative, with the ambitious goal of ensuring every person on Earth is protected by life-saving hazard alerts by 2027. ITU is proud to lead Pillar 3 of this initiative, which focuses on warning dissemination and communication. We know that a warning is only effective if it reaches those at risk in time for them to act, and this is where AI is proving to be a game-changer.
The journey from innovation to impact is being realized through the AI for EW4All sub-group, a powerful partnership convened by ITU alongside WMO, UNDRR, and IFRC, and joined by tech leaders like Microsoft, Google, and the GSMA . Together, we are moving from theoretical potential to practical, life-saving solutions.
One of our most significant innovations is the Early Warning Connectivity Map (EWCM). Developed with Microsoft, Planet Labs, and IHME, this AI-powered tool uses satellite imagery and connectivity data to create high-resolution population maps.
It answers a critical question: who can receive an alert, and who cannot? By identifying connectivity "cold spots," we help governments design targeted alerting strategies that reach the most vulnerable, bridging the digital divide rather than widening it.
The impact is measurable. In pilot countries like Fiji, Tonga, and Vanuatu, the EWCM has helped authorities identify major coverage gaps and strengthen their national strategies for last-mile inclusion. And by fusing complex datasets—from weather and radar to mobility patterns—AI models are improving forecast accuracy and extending lead times for tropical cyclone and flash-flood warnings by an average of 30 minutes. In disaster management, every minute saved is a life saved.
This progress is driven by a multi-stakeholder governance model that ensures ethical and equitable deployment. Through platforms like the AI for Good Global Summit, we are connecting innovators with governments, matchmaking donors with practitioners, and hosting Innovation Challenges to source cutting-edge AI solutions from around the world.
Our work is already influencing policy, with 15 governments adopting AI guidelines for disaster risk reduction under our technical assistance tracks.
Today, it is my pleasure to welcome you to the launch of the new report, “Leveraging AI to Enhance Multi-Hazard Warning Systems, “a joint publication by ITU, UNDRR, WMO, and IFRC. The report provides not only an overview but a practical resource and concrete examples of how AI is being applied. A key finding of the report is that AI is not a solution-in-isolation. It is rather, an enabler that depends on strong institutions, inclusivity and accessibility, quality data, resilient infrastructure, and above all, human expertise and partnerships.
In the area of warning dissemination and communication, breakthroughs include A—enabled dissemination platforms to expand warning coverage, voice-based alerts to reach non-literate populations, and AI-powered translation supporting multilingual public alerting, making early warning systems more inclusive and effective.
Looking Ahead …
As we look to the future, our focus remains on scaling these solutions. We are expanding the Connectivity Map to over 30 countries and building capacity so national authorities have the skills to interpret and apply this data. We are also integrating AI into National Emergency Telecommunication Plans, ensuring that no country is left behind.
The goal is clear: to transform AI innovation from a concept into a global public good that protects every community.
Thank you.
