GENEVA

The World Meteorological Congress on Friday endorsed new measures to integrate artificial intelligence (AI) into global weather forecasting and early warning systems, emphasizing that all countries should benefit from its life-saving potential.

The extraordinary congress, in a statement, called on the public, private, and academic sectors to collaborate on AI and machine learning technologies to improve predictions of hazards such as extreme heat and rainfall. It stressed that AI should complement, not replace, traditional forecasting methods.

"Early warnings are not an body abstraction. They give farmers the power to protect their crops and livestock. Enable families to evacuate safely. And protect entire communities from devastation," UN Secretary-General Antonio Guterres said, noting that countries with strong warning systems have six times lower disaster-related mortality.

WMO Secretary-General Celeste Saulo urged governments to turn "visibility into impact" and "recognition into investment," highlighting that global cooperation is vital to achieve the UN-backed Early Warnings for All initiative by 2027.

Congress resolutions included technical standards to anchor AI and data science in WMO's global observation and forecasting systems.

They also emphasized open data, ethical frameworks, and capacity building for developing nations, particularly least developed countries and small island states.

A pilot project between Norway and Malawi, showcasing AI-driven, Forecasts-in-a-Box, demonstrated how low-income countries can enhance prediction accuracy, the WMO said.