Governing the environmental impacts of AI
Over recent years, Artificial Intelligence (AI) has been promoted as a tool to tackle environmental challenges, from climate change to biodiversity loss. At the same time, the growth of AI itself produces significant environmental costs. Large data centres consume increasing amounts of energy and water, while the production of specialised hardware relies on critical raw materials that are often extracted under environmentally damaging and socially exploitative conditions.
Sarah Hartley, Aleksandra Stelmach, Emily Robinson, Ernesto Schwartz Marin, and Katie Ledingham are leading research exploring how these environmental impacts are experienced, understood, and governed. We convened an interdisciplinary workshop, Sustainability of AI, to map AI’s environmental costs and identify governance gaps. Our Legacy of AI project examines how researchers who use AI for environmental purposes negotiate responsibility and sustainability in their work. Lastly, our project on Environmentally Just AI examines AI’s environmental impacts through environmental social science investigating how governance gaps, supply chain inequalities, and public discourse shape the environmental consequences of AI.
