Wildlife Conservation Deployments Reveal AI Missing Elephants and Antelopes
Artificial intelligence is transforming wildlife research across Africa by processing hundreds of thousands of aerial photographs in hours instead of months, though the technology’s surging demand for electricity and water is drawing scrutiny on the eve of a major global climate summit. A study published on Nov. 7 by Reuters detailed how machine learning models deployed by the Kenyan conservation group Save the Elephants uncovered large populations of animals that traditional manual counts missed entirely.
During a large-scale aerial survey in 2025, researchers gathered approximately 200,000 images of a conservation area. Manual checks by nine specialists previously took nine months, but Nvidia graphics processing units now complete the task in a matter of hours.
The resulting counts surprised researchers. Frank Pope, head of Save the Elephants, stated that traditional censuses undercounted roughly a third of the elephants, half of the zebras and giraffes, and three-quarters of the oryx inhabiting arid African regions. Accurate population data determines conservation status, funding allocations, and habitat protection measures, making these algorithmic corrections vital for wildlife management.
Archival Digitization Accelerates Primate Research in Tanzania
In Tanzania’s Gombe National Park, the Jane Goodall Institute is applying similar technology to historical records. Researchers spent decades recording chimpanzee behavior by hand using the local Kicha language and specialized abbreviations. Processing this massive paper archive manually required between 13 and 15 years.
By collaborating with Amazon Web Services, the institute is now scanning, translating, and organizing decades of field notes in roughly two years, according to Liliana Pintea, Vice-President of the Institute for Environmental Research. These digitized records allow scientists to track long-term behavioral changes and assess how environmental shifts affect primate populations over time.
Beyond terrestrial animals, conservation groups are deploying automated systems in other ecosystems. Conservation International analyzes satellite data on temperature anomalies in Cambodia to detect forest fire hotspots and warn local patrols. Meanwhile, The Nature Conservancy places AI-enabled cameras on commercial fishing vessels to identify captured fish species on board, reducing the need to transmit heavy video files to remote servers.
Data Centers Demand More Power Ahead of COP31
Despite these conservation benefits, the rapid expansion of computing infrastructure creates new environmental strains. Powering servers and cooling data centers requires immense volumes of electricity and water. The International Energy Agency projects that global data center electricity consumption will reach 945 terawatt-hours annually by 2030, doubling 2024 levels and matching Japan’s yearly power use. Annual water consumption by these facilities is expected to climb from roughly 560 billion litres in 2023 to 1.2 trillion litres by the end of the decade.
Major environmental organizations are grappling with these trade-offs. The World Resources Institute limits resource-intensive tasks like image and video generation by reusing previous computational results. Greenpeace takes a more cautious approach, questioning the necessity of large-scale data center construction and urging careful evaluation of every new computing project.
These competing pressures form a central backdrop for the United Nations COP31 climate conference, scheduled to begin on Nov. 9 in Turkey. Organizers plan to introduce responsible technology development principles, while UN officials press major tech companies for greater transparency regarding energy use, water consumption, and greenhouse gas emissions.
Keep reading