Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Data centers: Friend or foe to sustainability?

What's at stake in the data center debate.

[Photo: Getty Images]

Profile photo of Saskia van Gendt

Chief Sustainability Officer · Blue Yonder

The expansion of artificial intelligence and the data center growth needed to run AI will be among the most transformative events in history. How we execute over the next five years will determine whether AI is a force for good. By adopting alternative energy and other environmental best practices while applying AI to solve sustainability challenges, we can turn AI growth from an environmental threat to a supercharger for sustainability.

THE ENVIRONMENTAL FOOTPRINT

Data centers are energy-hungry. Worldwide, they are estimated to have used 448 terawatt-hours of electricity in 2025. That’s almost as much as the 468 TWh of electricity used by France over the same period.

At this rate, the United Nations estimates that data center electricity consumption could surpass 945 TWh by 2030. The International Energy Agency estimates that carbon emissions attributable to data center energy use could increase by nearly 78%, from 180 million tonnes of CO2 today to 320 million tonnes by 2030 if we do not adopt cleaner energy sources at the same pace of growth. Fortunately, many of the hyperscalers—the huge cloud computing and data center companies—have committed to net-zero carbon emissions and are adopting renewable energy, including wind and solar. In 2025 most of the energy (67%) used in data centers came from renewables.

Coupled with clean energy sources, distributed battery storage can be a promising solution for powering data centers. It can store excess grid capacity, provide backup for intermittent renewables, and power data centers for up to the next decade.

Data centers also consume a lot of water, both indirectly through energy use and directly for cooling the servers. According to the U.N., in 2025 data centers used 4.5 trillion liters of water via electricity consumption. This is expected to rise to 9.3 trillion liters by 2030. According to a University of Georgia College of Agricultural & Environmental Sciences Field Report, a water-cooled data center can use 300,000 to 5 million gallons of water per day, stressing regions with limited supplies.

Immersion cooling could help address the water problem. Servers are submerged in special liquid coolants that are thermally (but not electrically) conductive. These coolants disperse heat 1,200 times better than air, are 40% to 50% more energy efficient, and reduce noise by eliminating fans and air-circulation equipment.

So, while the current trajectory of AI and data center growth points to increased energy use, carbon emissions, and associated impacts, real solutions exist if we continue to invest at scale.

INNOVATION AND AI SUPPLY CHAIN SUSTAINABILITY

The potential benefits from AI are considerable. AI can fuel innovation that improves our lives across industries, including agriculture, education, healthcare, manufacturing, science, software, and supply chains. Slowing data center expansion could jeopardize AI’s potential. The cost could come in many forms, including lower crop yields, delayed cancer diagnoses, or unchecked supply chain emissions.

Supply chains are key to sustainability efforts. Sixty percent of global carbon emissions come from supply chains, and 40% of these are cost-effectively addressable through existing solutions. It will take AI to help solve AI’s environmental challenges—and it is already having a huge impact on supply chains.

Machine learning is making demand forecasts and predictions more accurate and precise, helping align production with actual demand. This reduces the excess emissions and waste associated with the manufacturing, storage, logistics, and disposal of unnecessary products.

In logistics, AI is helping to optimize truckloads, identify backhaul opportunities, and optimize routes. This reduces emissions by minimizing empty miles and unnecessary trips.

Returns represent a huge opportunity for businesses to cut billions in waste and recapture lost revenue. AI is helping make the entire returns process more convenient, efficient, and sustainable with “smart dispositioning.” It identifies the most profitable and efficient way to route returns, so they can be resold at full value and are far less likely to become landfill.

AI helps companies become more sustainable by analyzing supply chain data, automating sustainability reporting, and identifying emissions “hotspots” and compliance issues across the organization, suppliers, and logistics providers. This data can be used at the point of decision-making to make more sustainable choices in procurement, logistics, and inventory planning.

TOWARD MORE SUSTAINABLE DATA CENTERS

Data centers have supply chains, too, making it critical to apply sustainable supply chain management to data centers themselves. Freight industry analysts estimate that every gigawatt of data center expansion in the U.S. will require 100,000 truckloads to move materials like cement, steel, and wiring. AI can help optimize the entire lifecycle of data center development, from semiconductor manufacturing to construction and operations. This reduces emissions and waste. Machine learning can help predict disruptions and recommend responses to avoid or mitigate the impact, making supply chains more resilient.  

Sustainable design also plays an important role here. Better building and environmental design can help alleviate some of the concerns that local populations have about data centers, making the buildings more appealing and energy-efficient, and reducing the noise they emit.

We should also consider the possibility that some data centers might become obsolete or no longer financially viable. Designing for disassembly enables the building to be easily reconfigured for other uses or completely disassembled, with the materials and parts reused in another building.

Ultimately, data centers do consume significant materials, energy, and water. But real solutions already exist to mitigate the footprint from data center growth. We need to continue adopting solutions and apply AI to solve sustainability challenges. With the right tools and strategies, we can decarbonize supply chains, build sustainable data centers, and unlock AI to help produce new life-enhancing innovations we could not have imagined.

Saskia van Gendt is chief sustainability officer of Blue Yonder.