Environmental NGOs are turning to artificial intelligence to boost their work, but they also worry about the extra electricity and water that AI data centres need. This article shows how they try to keep the benefits while limiting the costs.

Groups around the world now use AI to analyse climate data, monitor forest fires and count endangered species. These tools let them react faster and gather information that would be hard to collect by hand.

Conservation International has started using AI to examine satellite‑detected hotspots and send real‑time fire alerts to forest patrols in Cambodia. Save the Elephants in Kenya uses machine‑learning to turn a nine‑month, 200,000‑image survey into an analysis that takes only a few hours. The Jane Goodall Institute is digitising decades of chimpanzee field notes, cutting a project that could have taken 13‑15 years down to about two years.

The International Energy Agency estimates that by 2030 global data‑centre electricity use will reach about 945 terawatt‑hours a year and water consumption will rise to roughly 1.2 trillion litres, creating huge carbon emissions. UN Secretary‑General Antonio Guterres and industry leaders have warned that AI’s footprint must be transparent and sustainable, a point that will be discussed at the COP31 climate conference, urging companies to disclose the environmental impact of their AI systems.

The World Resources Institute limits AI impact by caching query results so identical searches are not re‑processed, and the Nature Conservancy runs AI on‑edge in digital cameras on fishing vessels, avoiding large data‑centre transfers. These steps show how NGOs can keep AI useful while cutting energy and water use.

Environmental groups are learning to balance AI’s powerful conservation benefits with the hidden cost of data‑centre resources. By using smarter computing methods, they hope to protect nature without adding to the climate problem.