Last year was the hottest year on record. Across the world, we are witnessing an increasing number of extreme events, from large-scale wildfires to floods and “hot tub” ocean temperatures, with devastating consequences for human life and our planet’s biodiversity.

The economic costs are already extremely high. Despite increasing environmental concern, my colleagues and I believe that much more can be done to save our planet. There is still much uncertainty in climate projections; gaps in both understanding of climate processes and computational capabilities limit the accuracy of climate predictions needed for optimal adaptation and resilience to climate change.

The use of AI helps narrow uncertainties in climate modeling, which can be used to optimize adaptation and potentially maximize equitable distribution of resources and inform actionable decisions. In fact, the application of AI to solve a range of climate and environmental problems is rapidly becoming widespread.

While we absolutely need to reduce greenhouse gas emissions to slow climate change, we also need accurate information about what will happen so we can better plan for and adapt to changes in the global climate system right now: Without changing current policies, we are on track to avoid more than 2°C – 3.5°F – of warming this century.

Columbia University, where I am a faculty member in the Fu Foundation School of Engineering and Applied Sciences and the School of Climate, in collaboration with the University at Albany, the State University of New York, and Esri, recently published a comprehensive “Artificial Intelligence Landscape.” Climate and Nature Assessment commissioned in collaboration with the Bezos Foundation Lands.

The report compares current AI applications and opportunities across various sectors of climate change mitigation, adaptation, and nature management, and is part of a companion package to the Bezos Earth Foundation’s $100 million AI for Climate and Nature Grand Challenge, which aims to use AI for climate change and conservation. The Nature Grand Challenge is a call to action to fund AI solutions for biodiversity conservation, power grid optimization, sustainable proteins, and other innovative ideas.

Since 2021, when I became co-director of the NSF Science and Technology Center at Columbia University, dedicated to improving climate predictions by combining physical modeling with artificial intelligence, I have been working with colleagues around the world to make our climate predictions more accurate by using artificial intelligence, and to provide a modern cloud-based platform for climate data.

The current generation of large AI models consumes an enormous amount of resources. If we hope that this new, cutting-edge technology will combat, rather than accelerate, climate change and the destruction of nature, our society must act quickly to encourage and fund cross-disciplinary collaborations to bring the capabilities of AI tools to the service of domain experts.

We must also rethink AI benchmarks to reflect these priorities and commit to data accessibility standards, encouraging open data, code, and models to drive algorithm and software development while protecting privacy.

Investing in education and workforce development to support AI professionals with cross-sector expertise is important. Our NSF-funded LEAP Center is already answering this call to action by providing the next generation of scientists with skills at the interface of climatology and AI, as well as a state-of-the-art platform for exploring climate data.

Many of our colleagues take a holistic approach to AI education, integrating fundamental and applied AI concepts into interdisciplinary curricula to prepare graduates for a rapidly changing world.

We must also ensure that individuals and communities around the world have equal access to the knowledge and information needed to use AI tools to ensure their current and future well-being in relation to climate change.

This is true locally by implementing solutions that can create impact and build a workforce in historically marginalized and disadvantaged communities most affected by climate change, and globally by investing in AI infrastructure, initiatives, and education in the Global South, where the majority of the next generation of this planet’s population is born.

Robust technology governance structures are also needed to ensure that AI is deployed ethically, sustainably, reliably, safely, and securely. Governing bodies around the world, such as the US, EU, and UN General Assembly, are already beginning to implement such policies. They must bolster government funding to support AI indefinitely.

Climate change is here. So is the age of AI. Achieving greater equality and prosperity on a thriving planet is still possible if we strategically apply our science, education, and innovation to achieve these goals.