AI for Good: Intelligence enhancing disaster resilience
Disasters don’t wait for models to mature. But artificial intelligence (AI) models trained to interact properly with people can provide foresight on a looming crisis, helping communities brace and protecting the local economy. They can also get emergency alerts out quickly and accurately, saving numerous lives.
The Global Initiative on Resilience to Natural Hazards through AI Solutions announced on 7 July its new AI Challenge for Drought Detection, focused on crowdsourcing better early-detection tools. The challenge for AI developers “will serve as a concrete example of translating research into something operational,” said Global Initiative Chair Monique Kuglitsch, Innovation Manager at Fraunhofer HHI.
Day 1 of the AI for Good Global Summit featured an in-depth discussion of AI in disaster resilience as a bridge between science, standards and innovation.
Speakers in the session considered how AI can enhance the work of public institutions and provide reliable, trustworthy services to vulnerable or disaster-affected populations, including hidden or underrepresented communities that could easily be overlooked.
Serge Wilmes, Luxembourg’s Minister of the Environment, Climate and Biodiversity and Minister for the Civil Service, framed resilience as “urgent and inherently horizontal.”
AI for disaster resilience is a multidisciplinary effort, and its impact depends on weaving science, standards, and implementation together as a unified effort, he said.
Urban AI solutions
Anacláudia Rossbach, Executive Director of UN-Habitat, underscored the need for responsible AI to strengthen sustainability in complex urban settings.
Combining satellite data with local institutional knowledge allows cities to react faster and plan more wisely for future scenarios.
“Cities increasingly use AI to protect, respond, and plan. But people, not just infrastructure, need to be at the centre of urban policy,” said Rossbach.
Urban AI solutions must also account for vulnerable people who tend to be often overlooked in the available data, she added.
Rossbach called for collaboration between universities and practitioners to establish interoperable urban AI standards and close the gap between theory and practice in AI-driven city management.
AI-human complementarity
AI is reshaping how weather forecasts are produced and analysed, too.
Celeste Saulo, Secretary-General of the World Meteorological Organization (WMO), noted that disaster management goes beyond predictions. AI can boost early warning systems, making them both timelier and more accessible.
But AI is only as strong as the data and observations underpinning it.
National meteorological services understand local risk in ways AI should enhance, not replace, Saulo added, suggesting that investments in AI for meteorology need to be channelled into country-level forecasting infrastructure.
Standards for integrating AI, if grounded in sound science, transparency, and equity, can drive cooperation to help the most vulnerable communities adopt, adapt, and sustain early warning systems, she said.
Where we are now
A decade ago, few would have believed in AI providing high-level of forecasting, added Florian Pappenberger, Director-General of the European Centre for Medium‑Range Weather Forecasts (ECMWF).
Along with continual improvements in accuracy, AI capabilities could result in a genuine democratization of forecasting, he suggested.
Rather than model size or compute power, the best indicator of success will be better decisions, fewer lives lost, and whether emergency services can act with greater confidence.
“AI will be remembered by how much it strengthened our collective resilience to disaster – nothing more, nothing less,” Pappenberger said.
Launch of the AI for Drought Detection Challenge at the AI for Good Global Summit 2026 with key partners
Learn more about the AI for Good Global Summit
Header image credit: Adobe Stock
