Everyone wants to talk about artificial intelligence. But according to Kamal Kishore, the UN Secretary-General's Special Representative for Disaster Risk Reduction, the bigger challenge may be something far less glamorous: collecting better data.
Speaking at the 2026 AI for Good Global Summit, Kishore argues that disaster data remains incomplete around the world. Even after major disasters, governments often lack basic information about who was affected and how communities experienced the crisis.
"Make the dull and boring task of collecting data really cool," says Kishore.
His call is simple: make collecting and organizing disaster data a priority. Without accurate, standardized information, AI-powered forecasting, risk analysis, and disaster response can only go so far.
As countries invest in the next generation of AI tools, Kishore says they must also invest in the fundamental work of building the datasets those systems rely on.
This conversation is presented by GZERO Media in partnership with Microsoft. The Global Stage series convenes global leaders for critical conversations on the geopolitical forces reshaping our world.
