Environmental groups balance AI benefits against data center costs

A tool for urgent conservation

Environmental organizations are increasingly turning to artificial intelligence to address the accelerating biodiversity and climate crises. These groups use the technology to analyze complex climate data, monitor forest fires in real-time, and perform large-scale wildlife population surveys. By leveraging AI, these organizations aim to move faster and make more effective use of the information they collect from the field.

For example, Conservation International recently began using AI to monitor satellite data for hotspots in Cambodia. The system sends immediate notifications to local forest units regarding possible fires in isolated and ecologically sensitive regions. This allows for a quicker response to threats that might otherwise go unnoticed by human observers alone.

The adoption of these tools is driven by a sense of urgency regarding the planet's health. Stuart Butchart, chief scientist at BirdLife International, noted that if they can save the planet more efficiently with limited resources by using AI, they have to engage with the possibility. He suggested that failing to use these tools could be seen as a dereliction of duty.

Accelerating scientific breakthroughs

The primary appeal of AI for these groups lies in its ability to process massive amounts of data that were previously impossible or too expensive to handle manually. In Kenya, the group Save the Elephants uses machine learning to study elephant communication and improve population surveys. They found that previous human surveys were missing significant portions of the local wildlife, including about a third of elephants and three-quarters of oryx populations.

The efficiency gains are substantial for researchers. An aerial survey of a protected area in Kenya produced roughly 200,000 images that would have taken nine people nine months to review by hand. Using AI models trained on Nvidia graphics platforms, the organization can now complete similar analyses in just a few hours. This speed helps make nature more measurable and, consequently, more attractive for conservation investment.

Similarly, the Jane Goodall Institute is using AI to digitize decades of handwritten field notes about chimpanzee behavior in Tanzania. These records were written in a local language and scientific shorthand, making them difficult to process. A tool developed with Amazon Web Services can now scan, translate, and structure these notes in minutes, cutting a projected 13-to-15-year project down to about two years.

The high cost of data centers

Despite the benefits, environmental groups are raising alarms about the physical footprint of the infrastructure required to power AI. Data centers are filled with servers that require vast amounts of electricity and water for cooling. The International Energy Agency estimates that global data centers will consume about 945 terawatt-hours of electricity annually by 2030, producing more than 300 million metric tons of carbon dioxide.

Water consumption is another growing concern for the industry. The International Energy Agency projects that global data center water use will rise from about 560 billion liters in 2023 to roughly 1.2 trillion liters annually by 2030. Furthermore, experts at the United Nations University Institute for Water, Environment and Health suggest that the total area occupied by data centers could more than double by 2030, reaching a size twice that of Jakarta.

Technology giants like Microsoft, Amazon, and Alphabet have reported rising emissions linked to their data center expansions. While Amazon claims its AI products are among the most sustainable due to low-carbon energy purchases, the overall trend creates a dilemma. Organizations are now forced to justify their use of AI by weighing its specific benefits against these significant environmental risks.

Strategies for sustainable AI

To mitigate these impacts, some organizations are adopting specific technical strategies to limit their carbon and water footprints. The World Resources Institute, for instance, limits resource-heavy applications like generating images or videos. They also use a technique called caching, which stores AI query results so that identical searches do not need to be reprocessed by the system.

The Nature Conservancy is taking a different approach by moving computation to the "edge." Instead of uploading massive amounts of video data from commercial fishing vessels to a centralized data center, they installed AI-powered digital cameras directly on the boats. This allows the information to be processed on-site, which significantly reduces the energy needed for data transfer and storage.

Other groups are focusing on policy and internal guidelines to ensure responsible use. Greenpeace International has imposed rules emphasizing transparency and human oversight, though they have expressed a preference to avoid the technology entirely due to its impacts. They are currently pushing back against the construction of massive new data centers, questioning whether such large-scale infrastructure is necessary in the first place.

Looking toward global policy

The tension between AI's utility and its environmental cost is expected to become a major topic at the UN's COP31 climate conference this month. Host Turkey will call on governments and businesses to sign a pledge outlining principles for sustainable AI development. The goal is to create a framework that ensures the technology is deployed in a way that does not undermine global climate goals.

UN Secretary-General Antonio Guterres is specifically calling on major AI companies to be more transparent about their environmental impact. He wants these corporations to publicly disclose how much energy and water their systems consume. This push for accountability comes as the industry faces increasing scrutiny over its ability to meet long-term sustainability commitments while expanding its infrastructure.

As these organizations navigate this contradiction, the focus is shifting toward intentionality. Groups like Conservation International emphasize that while AI has costs, it is a necessary tool for addressing urgent crises. The future of environmental activism will likely depend on finding a balance where the technology's ability to save the planet is not outweighed by the resources required to run it.