...
By Suzanne Blake
Reporter
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
Artificial intelligence is fueling a boom in data centers as well as concern about their environmental footprint.
Data facilities consume significant electricity and water, especially as demand for tools like generative AI grows. Altogether, U.S. data center annual energy use in 2023 was recorded at 176 terawatt-hours or approximately 4.4 percent of U.S. annual electricity consumption, according to the Lawrence Berkeley National Laboratory.
But amid the environmental alarm, it can be useful to put that impact in context. Compared with many major industries and everyday activities, the environmental footprint of AI data centers is still relatively small, experts say.
Read More on Tech
“This data center dependency isn’t always portrayed fairly,” Alex Beene, a financial literacy instructor for the University of Tennessee at Martin, told Newsweek. “When compared to other economic areas like air travel and manufacturing, some data centers actually consume less energy and could over time be used to actually assist in reducing emissions.”
Why It Matters
Data centers globally account for roughly 1 to 2 percent of electricity use, and this share is expected to rise with AI expansion. But many of the largest drivers of greenhouse gas emissions, such as fossil fuels, agriculture and heavy industry, remain far more significant contributors to climate change today.
This can clarify where the biggest environmental pressures really lie and where policy or consumer behavior could have the greatest impact.
What To Know
AI data centers affect the environment through energy use, water consumption for cooling, and construction emissions.
However, their footprint varies widely depending on energy sources and whether these are renewable or fossil fuels.
“Silicon Valley has done an excellent job selling the promise of AI, but we're still waiting to see whether the benefits justify the growing costs,” Kevin Thompson, the CEO of 9i Capital Group and the host of the 9innings podcast, told Newsweek. “AI data centers consume enormous amounts of electricity and water, in some cases rivaling the usage of entire communities.”
Newsweek compiled several activities and industries that consistently have larger overall environmental impacts than AI data centers.
Beef Production
Beef is widely considered one of the most environmentally intensive foods. Cattle produce large amounts of methane, a potent greenhouse gas, and require extensive land and water.
...
Globally, livestock emissions exceed those of the entire aviation sector, with beef the single largest contributor within agriculture.
“Beef production, manufacturing, and energy generation all consume significant resources, but their benefits are tangible and easy to identify,” Thompson said. “With AI, much of the value remains theoretical for the average American, while the environmental costs are already measurable.”
Coal-Fired Power Plants
Today, coal is still one of the dirtiest energy sources. A single large coal plant can emit millions of tons of carbon dioxide annually, often surpassing emissions from many individual data center campuses.
However, AI data centers powered by coal-heavy grids also have a much larger footprint than those running on renewable energy.
Cement Production
Cement manufacturing accounts for about 7 to 8 percent of global carbon dioxide emissions, driven by both fuel use and chemical reactions during production.
That makes it one of the largest industrial contributors to climate change and far surpasses emissions linked specifically to AI computing.
Commercial Aviation
Air travel contributes hundreds of millions of tons of carbon dioxide yearly, with long-haul flights generating several tons per passenger.
Even as data center demand rises, aviation’s total climate impact remains significantly larger.
Still, Thompson cautioned that there are larger societal risks to an increased reliance on AI that its environmental impact doesn’t measure.
“Beyond economics, there is a cultural concern,” Thompson said. “More people are relying on AI to think, write, and research for them. At some point, we have to ask whether we're training the Ai or whether the Ai is training us.”
Fast Fashion
The fashion industry generates substantial emissions through manufacturing, shipping and disposal, along with water pollution and microplastics.
Fashion production is responsible for around 10 percent of global carbon emissions, according to Business Insider, and 85 percent of textiles are thrown away each year, alongside 500,000 tons of microfibers released into the ocean. That’s roughly the same as 50 billion plastic bottles, according to Earth.org.
Home Heating and Air Conditioning
Heating and cooling buildings account for a major share of energy consumption, especially in countries like the United States.
Because nearly every household relies on climate control, the total emissions from buildings far exceed those from data centers alone.
Deforestation
Clearing forests releases stored carbon and can also destroy biodiversity and disrupt ecosystems in the process.
On a global scale, deforestation contributes more to climate change than AI infrastructure and also has long-term environmental consequences beyond emissions.
Cryptocurrency Mining
At its peak, Bitcoin mining consumed electricity on the scale of entire countries.
“Cryptocurrency mining is an extremely energy-intensive process that threatens the ability of governments across the globe to reduce our dependence on climate-warming fossil fuels,” Earthjustice said in a 2022 report.
In 2021 to 2022, Bitcoin consumed an estimated 36 billion kilowatt-hours of electricity. That was as much as the electricity consumed in Maine, New Hampshire, Vermont, and Rhode Island put together. However, other reports find the energy usage to be much higher.
Cruise Ships
Cruise ships burn large amounts of fuel and generate pollution and wastewater, often with a high environmental cost per traveler compared with other forms of transportation.
AI data centers are undeniably energy-hungry, experts say, but when measured against other global environmental challenges today, industries like agriculture, manufacturing and transportation continue to have much larger overall footprints.
However, that doesn’t mean AI’s hit to the environment should fall on deaf ears, said Drew Powers, the founder of Illinois-based Powers Financial Group.
“The problem with this type of ‘better or worse’ thinking is it’s a false equivalency. By this rationale, anything can be justified if you can find only one thing in the entire world that is worse,” Powers told Newsweek. “The fact of the matter is AI uses a lot of natural resources to power, and that should be concerning to everyone.”
What Happens Next
The environmental impact of AI is expected to continue to grow rapidly, increasing concerns from regulators and environmental researchers.
“There are still very real concerns over larger facilities in some states not being beneficial to those in the immediate area, and it’s up to local and state governments to protect the people and resources in these communities,” Beene said.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines