WASHINGTON, UNITED STATES / MANILA, PHILIPPINES — November 10, 2026 — Artificial intelligence is becoming an unlikely ally in the fight against climate change and wildlife extinction. But the same technology helping scientists protect endangered species and monitor threatened forests is also fueling a massive expansion of electricity-hungry, water-consuming data centers.
From the forests of Cambodia to the wildlife reserves of Kenya, environmental organizations are using AI to detect potential wildfires, count endangered animals and analyze research that once required years of manual work.
Yet behind these breakthroughs lies an increasingly uncomfortable contradiction.
The International Energy Agency (IEA) estimates that global data centers could consume approximately 945 terawatt-hours of electricity annually by 2030, more than double their estimated consumption in 2024.
Meanwhile, worldwide data-center water consumption could reach approximately 1.2 trillion liters annually by 2030, according to projections cited in Reuters’ investigation.
The implications are raising difficult questions for environmental organizations that have spent decades campaigning against pollution, fossil-fuel emissions and the overuse of natural resources.
In interviews with Reuters, 15 major environmental, conservation and development organizations confirmed that they were using artificial intelligence in some form.
All 15 also expressed concern about its environmental costs.
The findings expose a growing dilemma for the global conservation movement: Can environmental groups justify using energy-intensive AI systems when those same systems contribute to the pressures they are trying to reduce?
Environmental Groups Turn to AI Despite Growing Climate Concerns
Artificial intelligence is changing the way scientists collect and interpret environmental information.
Conservation teams have traditionally relied on field surveys, satellite observations, photographs, handwritten records and extensive manual analysis.
These methods produce valuable scientific information, but processing large datasets can require substantial time and personnel.
Machine-learning tools can accelerate some of that work.
For example, AI can help identify animals in aerial photographs, flag potential wildfire threats and categorize information gathered through environmental monitoring.
These capabilities can be particularly valuable for organizations operating with limited budgets.
However, using AI also requires computing resources.
Complex models often run on powerful processors housed in data centers that consume substantial electricity and may require water for cooling.
The environmental impact depends on the electricity supply, equipment efficiency, cooling technology, hardware production and the location of the computing facilities.
The result is a conflict between the environmental benefits of particular AI applications and the broader footprint of the infrastructure supporting them.
AI Helps Detect Wildfires in Cambodia Before They Spread
One example comes from Conservation International, a nonprofit environmental organization working to protect biodiversity and natural ecosystems.
The group has begun using artificial intelligence to analyze potential wildfire hotspots identified through satellite observations in Cambodia.
The system helps identify areas that may require closer attention and sends alerts to community forest patrol teams.
For remote forests, early information can be critical.
Wildfires can damage ecosystems, destroy vegetation and threaten wildlife habitats.
Detecting suspicious hotspots earlier may help patrol teams prioritize where to investigate.
The technology does not eliminate the need for local conservation workers.
Instead, it helps them make faster decisions using information that would otherwise require more extensive analysis.
Conservation International’s AI lead, Ali Swanson, acknowledged the environmental costs of the technology while emphasizing the need for conservation organizations to respond more quickly to climate and biodiversity threats.
The Cambodian project illustrates one argument in favor of responsible AI adoption: using limited computing resources for a clearly defined environmental purpose may deliver benefits that justify those resources.
But demonstrating those benefits requires monitoring both the system’s accuracy and its environmental footprint.
Kenya’s Elephant Researchers Discover What Human Surveys Were Missing
One of the most striking examples of AI’s potential comes from Kenya.
Save the Elephants, a conservation organization working with the Wildlife Conservation Network, is using machine learning to improve wildlife population surveys and study elephant communication.
In a 2017 aerial survey of a large protected area in Kenya, researchers collected approximately 200,000 photographs.
Reviewing those images manually required nine people working for nine months.
AI systems trained using Nvidia GeForce graphics processors can now complete comparable image analysis in hours, according to the organization.
The results revealed significant differences between traditional manual surveys and AI-assisted counts.
Researchers reported that manual surveys had missed approximately:
- One-third of elephants.
- Half of zebras and giraffes.
- Three-quarters of oryx populations.
These figures describe findings from the organization’s survey comparisons. They should not be interpreted as universal error rates across all wildlife surveys.
Nevertheless, the results suggest that AI-assisted analysis can help researchers build a more accurate understanding of animal populations.
That information matters because conservation decisions often depend on population estimates.
Underestimating the number of animals in an ecosystem can distort scientific assessments and influence funding priorities.
More accurate data could also help governments, conservation organizations and donors decide where protection efforts are needed.
Save the Elephants chief executive Frank Pope believes improved wildlife measurement could strengthen the case for biodiversity investment.
But while AI may help conservationists process their data more efficiently, its success still depends on the quality of images, model training and human validation.
Jane Goodall’s Chimpanzee Research Could Be Digitized Years Ahead of Schedule
Another major conservation breakthrough involves the research legacy of the late primatologist Jane Goodall.
For decades, researchers recorded observations of chimpanzee behavior in Tanzania’s Gombe National Park.
Many records were handwritten and contained scientific shorthand alongside the local Kiha language.
Turning this enormous archive into searchable digital information was expected to require approximately 13 to 15 years.
The Jane Goodall Institute is now using AI technology developed with Amazon Web Services to help scan, translate and structure those historical records.
According to Lilian Pintea, the institute’s vice president of conservation science, the revised project could take approximately two years.
That represents a substantial reduction in the time required to make the material more accessible.
The archive could help scientists examine long-term changes in chimpanzee behavior, habitat conditions and relationships with the surrounding ecosystem.
The technology also supports the preservation of scientific observations collected over generations.
However, digitization does not automatically guarantee perfect transcription or translation.
Historical research contains details that may require specialist interpretation.
Human researchers remain responsible for validating results and preserving the meaning of the original observations.
The institute’s position emphasizes the importance of using AI to support scientific expertise rather than replace it.
The Environmental Price: Data Center Electricity Demand Could Double by 2030
The conservation benefits are substantial, but the growing infrastructure required to power artificial intelligence presents its own environmental challenge.
According to the International Energy Agency’s Energy and AI report, global data centers consumed an estimated 415 terawatt-hours of electricity in 2024.
That represented approximately 1.5% of global electricity consumption.
Under the IEA’s 2025 base-case projection, demand could rise to approximately 945 terawatt-hours by 2030.
A subsequent IEA update placed its central forecast at around 950 terawatt-hours, reinforcing the expectation that demand could roughly double.
For context, the projected figure is comparable to Japan’s annual electricity consumption.
Data centers could account for approximately 3% of global electricity demand by the end of the decade.
Artificial intelligence is an important driver of this growth, particularly because of the expansion of high-performance computing systems.
However, the figures include all data-center workloads, such as cloud computing, conventional digital services, networking and data storage.
It would therefore be inaccurate to attribute the entire projected electricity requirement to AI alone.
Even at a relatively small share of global demand, data centers can place considerable pressure on local electricity systems when many facilities are concentrated in one region.
The consequences depend on the availability of generation capacity, transmission infrastructure and the mix of energy sources used to supply them.
Could Data Centers Use 1.2 Trillion Liters of Water Every Year?
Electricity consumption is only part of the problem.
Some large data centers also require substantial water resources to manage heat generated by servers and other equipment.
Water may be used directly in cooling systems, while electricity generation can create additional water-related impacts.
According to projections cited by Reuters, global data-center water consumption could increase from approximately 560 billion liters in 2023 to about 1.2 trillion liters annually by 2030.
That increase would place additional scrutiny on the sector’s resource requirements.
However, water consumption differs significantly by facility.
Factors include local climate, cooling design, electricity source, operating efficiency and the distinction between water withdrawn and water actually consumed.
Some data centers use evaporative cooling systems.
Others rely on air cooling or closed-loop technologies designed to reduce direct water consumption.
The environmental consequences also depend on local water availability.
A facility in a water-rich region may face different challenges from one built in a drought-prone area.
The global projections therefore do not establish that every data center consumes excessive amounts of water or that a particular facility is causing local shortages.
But they underscore why water-use reporting is becoming increasingly important.
For communities experiencing drought and water-security problems, the rapid expansion of digital infrastructure raises questions about how scarce resources should be allocated.
Greenpeace Questions Whether the AI Boom Really Needs So Many Data Centers
Not every environmental organization believes the solution is simply to use AI more efficiently.
Greenpeace International is taking a more critical position.
The organization acknowledges that AI features are already incorporated into some software platforms it uses.
However, it is developing internal guidelines emphasizing intentional use, transparency and human oversight.
Greenpeace also questions the scale of the global data-center expansion.
Its argument is that policymakers and technology companies should examine whether all proposed AI infrastructure is necessary before approving extensive construction.
This differs from the approach taken by organizations that see AI as an increasingly essential conservation tool.
Greenpeace’s concerns reflect the broader environmental risks associated with expanding electricity generation, hardware manufacturing, construction and water consumption.
Large data centers may require new power infrastructure, additional grid capacity and substantial investment in cooling equipment.
Even when electricity is purchased from renewable sources, manufacturing and construction can generate emissions.
The debate therefore extends beyond the efficiency of individual AI models.
It concerns the overall scale of computing demand and whether the environmental benefits justify the resources required.
The growing public resistance to data-center development has already affected investment plans in multiple regions.
As previously examined by WWC ONE MEDIA in its report on the spreading AI data-center backlash across Europe, Asia and Africa, communities and regulators are increasingly questioning large digital infrastructure projects.
Concerns include electricity costs, water availability, land use and the burden placed on existing infrastructure.
The challenge for governments is balancing technological competitiveness and investment with environmental protection and community interests.
World Resources Institute Is Finding Ways to Reduce AI’s Environmental Footprint
Some environmental organizations are responding by changing how they use artificial intelligence.
The World Resources Institute, which applies AI to climate and environmental analysis, has adopted measures intended to reduce unnecessary computing.
These include limiting resource-intensive activities such as certain image and video generation tasks.
The organization also reuses previously generated results when possible, reducing the need to process identical requests repeatedly.
This practice, known as caching, can lower computing requirements for suitable workloads.
The approach reflects an important principle: not every task requires the most sophisticated or resource-intensive AI system.
Smaller models, conventional software or simpler analytical methods may be sufficient for some applications.
Choosing the least resource-intensive method capable of completing a task can help organizations limit unnecessary demand.
However, the environmental savings depend on implementation.
Efficiency improvements do not always result in lower total consumption if overall usage continues to increase rapidly.
Organizations must therefore consider both per-task efficiency and the total amount of computing they use.
The Nature Conservancy Brings AI Directly Onto Fishing Vessels
The Nature Conservancy is testing another approach to reducing the resource requirements of AI-assisted environmental monitoring.
The organization analyzes video recorded aboard commercial fishing vessels to help identify the types and numbers of fish being caught.
Traditionally, processing large quantities of footage through centralized systems could involve substantial data transfers.
Instead, The Nature Conservancy has deployed AI-powered cameras that process information directly aboard fishing vessels.
The approach is called edge computing.
By analyzing information closer to where it is collected, the organization can reduce the need to transmit extensive video files to remote data centers.
This can improve efficiency and reduce certain communications and computing requirements.
It can also make monitoring more practical in locations with limited connectivity.
However, edge computing is not automatically more energy-efficient for every task.
Its environmental benefits depend on the hardware, software, workload and systems it replaces.
The Nature Conservancy’s approach nevertheless demonstrates that conservation organizations have options beyond sending every dataset to a large centralized computing facility.
Big Tech Faces Rising Pressure Over Emissions
The environmental debate is also becoming more significant for major technology companies.
Microsoft, Amazon and Google’s parent company Alphabet have reported emissions-related pressures associated partly with expanding data-center infrastructure.
Artificial intelligence is contributing to demand for powerful computing systems, new facilities and additional electricity supplies.
The resulting expansion creates challenges for corporate climate commitments.
Amazon told Reuters that its use of low-carbon electricity and efforts to improve efficiency made its AI offerings comparatively sustainable.
That is the company’s own assessment and should not be presented as an independently established industry ranking.
Reuters reported that Microsoft and Google did not respond to requests for comment for its investigation.
The wider concern is that more efficient technology may still require more total electricity if demand expands quickly.
A data center can reduce energy use per computation while its total power consumption increases as it serves more customers.
This makes absolute emissions, total electricity consumption and water use important alongside efficiency indicators.
Independent measurement and transparent reporting will be essential in determining whether the sector’s climate commitments are being fulfilled.
COP31 Could Put Sustainable AI Under a Global Spotlight
The debate comes as governments meet in Antalya, Türkiye, for the COP31 United Nations climate conference.
The summit is taking place in November 2026, with the environmental impact of artificial intelligence expected to feature in discussions about technology and sustainability.
Türkiye is encouraging governments, companies and other organizations to support principles for more sustainable AI development and deployment.
United Nations Secretary-General António Guterres has also urged major artificial-intelligence companies to disclose more information about the environmental footprint of their systems.
The pressure reflects a growing concern that the rapid expansion of computing infrastructure is outpacing environmental reporting and oversight.
For policymakers, the immediate challenge is obtaining accurate information.
Without reliable data about energy use, water requirements and emissions, it becomes difficult to assess whether an AI application delivers a net environmental benefit.
However, voluntary commitments discussed at a climate summit are not equivalent to legally binding international regulations.
Their impact will depend on whether governments and businesses adopt measurable standards and implement them.
Philippine AI Expansion Makes the Debate Relevant at Home
The Philippines is also entering a period of growing investment in digital infrastructure.
Telecommunications companies, power producers and technology developers are exploring additional data-center capacity to support artificial intelligence, cloud services and enterprise computing.
But the expansion raises questions about electricity supply, operating costs and environmental sustainability.
One example involves Meralco PowerGen Corporation and VITRO, which are studying potential data-center projects supported by renewable energy.
As WWC ONE MEDIA reported on MGEN and VITRO’s renewable-powered data-center initiative, the companies are assessing how power generation and digital infrastructure might be developed together.
The proposal remains at the feasibility-study stage.
It does not establish that the facilities have been constructed or that their environmental impacts have already been resolved.
Nevertheless, the initiative highlights an emerging commercial challenge.
Data-center developers need reliable, competitively priced electricity, while energy providers are searching for new customers as digital demand grows.
The environmental outcome will depend on the final project designs, electricity supply arrangements and operating practices.
Converge’s AI-Ready Data Center Adds Another Dimension
Converge ICT Solutions is also expanding its digital infrastructure to support cloud computing and AI-related workloads.
Its Angeles Data Center in Pampanga forms part of the company’s broader efforts to strengthen domestic data hosting and connectivity.
The facility has been designed for AI-ready and high-performance computing requirements.
In related coverage, WWC ONE MEDIA examined Converge’s domestic data infrastructure expansion and its implications for Philippine AI and cloud services.
The company’s investments reflect increasing demand for local processing capacity and data-storage services.
For the Philippines, the opportunity includes attracting technology investment and supporting domestic businesses.
But additional computing infrastructure can also increase demand for electricity, grid upgrades and cooling.
These requirements make coordination between energy providers, developers and regulators particularly important.
The long-term objective should be to expand digital services without creating unacceptable pressure on local resources or infrastructure.
Could AI Data Centers Increase Electricity Costs for Ordinary Consumers?
One concern receiving growing attention is whether households could end up paying some of the infrastructure costs associated with large data-center developments.
Massive computing facilities can require new substations, transmission equipment and additional electricity generation.
Depending on the regulatory system and how those projects are financed, some costs may be recovered through electricity tariffs.
That possibility is prompting policy debates in the United States.
As WWC ONE MEDIA reported in its coverage of congressional action on AI data-center electricity costs, US lawmakers have examined measures intended to prevent the financial burden of large digital infrastructure investments from being unfairly shifted onto ordinary consumers.
The issue is relevant to other countries expanding data-center capacity, including the Philippines.
However, the construction of a data center does not automatically mean residential electricity rates will increase.
The actual impact depends on electricity markets, grid capacity, regulatory decisions, financing arrangements and whether new infrastructure is paid for by developers or other customers.
For policymakers, transparent planning can help determine who benefits from large projects and who should pay for the infrastructure required to support them.
AI Could Help Solve Environmental Problems — But It Cannot Replace Environmental Policy
The achievements reported by conservation organizations demonstrate that artificial intelligence can provide valuable scientific tools.
Machine learning can process wildlife images, analyze environmental observations and help identify emerging threats.
But AI is not a substitute for protecting forests, reducing emissions, conserving water or enforcing environmental regulations.
A system that identifies an endangered animal cannot guarantee that its habitat will be protected.
A wildfire detection model cannot prevent fires unless warnings lead to effective action.
And a more accurate wildlife survey cannot restore an ecosystem without the necessary conservation policies and resources.
The real value of AI depends on how its findings are used.
That is why the technology should be judged through measurable environmental outcomes rather than the sophistication of the computing systems involved.
The most sustainable approach may involve combining AI with conventional monitoring, scientific expertise, local knowledge and strong environmental governance.
What Needs to Happen Next?
The growing relationship between artificial intelligence and conservation creates several priorities for governments, technology developers and environmental groups.
First, improve environmental transparency. Technology companies should disclose credible information about electricity consumption, water use and greenhouse-gas emissions.
Second, match computing resources to actual needs. Organizations should avoid unnecessarily intensive AI applications when simpler methods can deliver the same results.
Third, prioritize sustainable infrastructure. Data-center developers should assess electricity sources, cooling requirements, grid impacts and local water availability before construction.
Fourth, verify conservation benefits. AI systems should be evaluated against practical results, including improved wildlife monitoring, faster environmental responses and more effective protection measures.
Fifth, protect affected communities. Infrastructure expansion should consider the interests of residents, electricity consumers and communities exposed to water shortages or environmental pressures.
These priorities are not mutually exclusive.
It is possible to support beneficial AI research while demanding stronger environmental safeguards from the computing industry.
The Bigger Picture: Can Artificial Intelligence Save the Planet Without Damaging It?
Artificial intelligence is giving conservation organizations capabilities that would have been difficult to imagine only a few years ago.
Researchers in Kenya can analyze wildlife images in hours instead of months.
Conservation teams in Cambodia can respond more quickly to potential forest-fire threats.
The Jane Goodall Institute expects to digitize decades of chimpanzee research years ahead of its original schedule.
These developments demonstrate the potential for technology to improve environmental protection.
But the infrastructure supporting AI is also expanding rapidly.
Global data-center electricity consumption could roughly double by 2030.
Water use could increase substantially, while additional facilities create demand for land, hardware and electricity-generation infrastructure.
For environmental organizations, the answer may not be to reject artificial intelligence entirely.
Instead, it may be to use it selectively, measure its consequences and demand accountability from the technology companies providing it.
For governments, the challenge is ensuring that rapid digital expansion does not compromise environmental and community interests.
And for the global AI industry, the ultimate test will be whether technological progress can be separated from unchecked resource consumption.
Artificial intelligence is helping scientists discover overlooked wildlife, protect threatened forests and unlock decades of environmental research. But as the data centers powering these breakthroughs demand ever more electricity and water, could the technology being used to save nature eventually become one of the biggest new threats to the planet it promises to protect?
