Our health, both physical and mental, is one of the principal ways we experience the impacts of weather and climate. As heatwaves, drought and flooding intensify, so does their toll on health. Yet despite these growing risks, weather and climate information remains insufficiently integrated into routine health decision-making. In many countries, health leaders making decisions about staffing, supplies, preparedness, or public health interventions still do so without access to tailored climate and weather intelligence. As our climate continues to change the frequency and severity of many health-related hazards, closing this gap becomes ever more urgent.
For decades, the challenge has not only been scientific but operational. Developing and maintaining forecasting systems capable of supporting specific health decisions has required significant investments in infrastructure, expertise, observations, and computing power. Many National Meteorological and Hydrological Services, particularly in developing countries, have relied on adapting forecasting systems developed elsewhere, often with limited ability to tailor them to local contexts and health priorities. As a result, climate and weather information has too often remained peripheral to health planning rather than becoming an integral part of it.
Artificial intelligence is beginning to change this equation. AI-enabled prediction systems offer the prospect of producing forecasts more quickly, more efficiently, and at a scale that was previously unimaginable. They are lowering technical barriers and creating new opportunities for countries to develop forecasting services tailored to their own needs. Importantly, they open the possibility of designing prediction systems not only around meteorological variables, but around the health decisions they are intended to inform.
This potential is significant. It offers an opportunity to accelerate what WMO refers to as climate-health intelligence: the systematic integration of weather, climate, environmental, and health information into decision-making processes that protect lives and strengthen resilience. By helping reveal relationships between climate risks and health outcomes, AI may enable earlier interventions, more targeted preparedness measures, and more effective use of limited health resources.
However, technological capability alone will not translate into better outcomes. The deployment of AI in climate and health forecasting raises equally important questions of governance, responsibility, and trust. Health systems do not act on information simply because it is available; they act on information they trust. Trust arises not only from scientific performance, but from clear institutional mandates, transparent methodologies, and well-defined accountability across the prediction-to-action chain. In most countries, these responsibilities are shared across National Meteorological and Hydrological Services, Ministries of Health, public health authorities, research institutions, and emergency management agencies. Strengthening the governance arrangements that connect these actors will be as important as advancing the technology itself.
This report recognizes that the future of climate-health forecasting depends on both innovation and stewardship. It examines how emerging AI capabilities can support more decision-relevant forecasting while highlighting the institutional foundations required for their responsible use. Success should not be measured solely by improvements in forecast accuracy, but by whether decision-makers receive information they can act upon confidently and effectively.
AI has made forecasting systems faster to build, easier to customize, and more accessible than ever before. Yet it has not answered the fundamental questions of governance, accountability, and public trust that accompany their adoption. Those questions must be addressed now, while the next generation of climate-health services is being shaped.
Véronique Bouchet
Senior Director, Department of Science, Services and Capacity Development,
World Meteorological Organization