11/17/2026 / By Chase Codewell

The carbon footprint produced by U.S. hyperscale data centers rivals that of the European country of Denmark, according to a new study published in Nature Sustainability on Monday, Nov. 16 [1].
“The reported emissions are a floor estimate,” said Francesca Dominici, the paper’s corresponding author and a professor of biostatistics at Harvard University. She added that the true figure is already likely to be several times higher given the number of operating and planned hyperscale facilities.
The findings arrive as the rapid expansion of artificial intelligence (AI) infrastructure has become a focus of public debate across the United States. The paper was published as United Nations climate talks in Turkey introduced a voluntary pledge addressing artificial intelligence’s role in both climate solutions and environmental risks.
Researchers identified 675 sites meeting their definition of hyperscale data centers but excluded hundreds because of difficulty obtaining good data, according to the paper. There are now approximately 1,200 such facilities operating across the United States, with hundreds more planned, meaning the true electricity and emissions figures are likely several times higher, the authors said.
The midpoint electrical power in the sample was 40 megawatts. Capacity is measured in megawatts because the electricity needed for computing and cooling is the main constraint, the report stated.
The team identified the mix of energy sources powering each data center, allowing them to quantify emissions, according to the paper. Data centers in Virginia, Oregon, Ohio and Iowa accounted for more than half of the electricity in the study, according to the research [2].
Dominici said the work was motivated by a lack of peer-reviewed study on the topic since 2018, even as AI and climate change have become major talking points. She called for stricter reporting requirements on data centers so that researchers can draw more precise conclusions. “We need to figure out the path forward to build these data centers in a responsible manner,” Dominici said.
Emissions across the facilities ranged from 21 to 31 million metric tons of carbon dioxide, according to the study [2]. Some 54% of the electricity was attributed to fossil fuels, 25% to renewables and 21% to nuclear, according to the paper.
Researchers noted that U.S. data center average carbon intensity was significantly higher than the carbon intensity of several European countries, including France, Spain, Italy, the United Kingdom and Germany [3]. The new estimate for cumulative carbon emissions is two to three times larger than previous estimates, according to reporting on the study [4].
Separate reporting has projected that data centers could account for 9% to 17% of national electricity load by 2030, up from 4% to 5% today [5]. Harvard’s numbers exceed many earlier estimates by two to three times, according to RaillyNews [6].
Gina McCarthy, who served as White House national climate adviser under former U.S. President Joe Biden, said in response to the study: “Sadly, it doesn’t surprise me that the majority of hyperscale data centers rely on fossil fuels for their power generation despite the profound and lasting climate and health impacts.” McCarthy added the study was an “important step” in giving communities and policymakers the reliable, independent data they need to inform decisions.
The paper comes as UN climate chief Simon Stiell highlighted the “great potential and great peril” of artificial intelligence in the battle against global warming during the COP31 talks in Turkey. The voluntary “Antalya Artificial Intelligence Pledge” seeks to tap the technology’s potential in finding climate solutions and assess its growing environmental risks, according to officials. The study was published as the AI boom has drawn scrutiny from health researchers as well.
A report published by Children’s Health Defense states that air pollution from artificial intelligence data centers may affect children’s developing brains and lungs [7]. The report focuses on data centers in the U.S. that power AI systems and require large amounts of electricity, the report states [7]. Nearly 1,900 public schools in the U.S. are within one mile of an AI data center, prompting health concerns for children, according to researchers [8].
There are signs of a possible slowdown. In the 2026 midterm elections, voters in dozens of American cities approved restrictions or outright bans on data centers, according to election results.
A June Yale Program on Climate Communication poll cited by CNBC found that 58% of registered voters oppose data center construction in their own area, including 53% of conservative Republicans and 74% of liberal Democrats [9]. Data center projects face unprecedented local opposition across the U.S., with 75 projects worth $130 billion delayed or blocked in Q1 2026 alone, according to reporting [10].
Progressive lawmakers Sen. Bernie Sanders (I-VT) and Rep. Alexandria Ocasio-Cortez (D-NY) have proposed a national moratorium, with conservative Sen. Josh Hawley (R-MO) endorsing the idea, according to statements from the lawmakers. New York Gov. Kathy Hochul (D) signed an executive order on July 14, 2026, ordering a statewide moratorium on large data centers, making it the first state in the country to suspend construction of such facilities, according to officials [11]. The order halts the construction and expansion of data centers for up to one year while state regulations are reviewed, officials said [11].
Critics of climate regulations said restrictions on domestic energy production can harm competitiveness in energy-intensive areas such as AI and data centers, and some analysts argue that carbon dioxide benefits plants and that climate narratives are misleading. Industry representatives said data centers require reliable power and that policy should balance environmental concerns with economic and technological needs, according to public comments.
Despite acknowledging that “AI is here to stay,” Dominici pointed out the need “to build these data centers in a responsible manner.” Separate research has noted that the power infrastructure required to serve data centers could need three to four times more metal than the data centers they support, according to a report from Wood Mackenzie [12].
Data centers could consume about 18 billion cubic feet of natural gas per day by 2035, nearly double an earlier forecast, according to BloombergNEF [13]. Google saw its carbon emissions rise 18% in 2025 compared to 2024 as it increasingly uses natural gas to power its data centers, according to Politico [14].
Dominici called for stricter reporting requirements on data centers so that researchers can draw more precise conclusions. The paper’s authors said their findings reframe large data centers as significant stationary emission sources rather than minor add-ons to the energy system [6].

Tagged Under:
artificial intelligence, carbon dioxide, carbon footprint, Censored Science, computing, data centers, electricity, energy consumption, environ, fossil fuel, future tech, Glitch, hyperscale, information technology, power, power grid, research, robots
This article may contain statements that reflect the opinion of the author
COPYRIGHT © 2017 GLITCH.NEWS
All content posted on this site is protected under Free Speech. Glitch.news is not responsible for content written by contributing authors. The information on this site is provided for educational and entertainment purposes only. It is not intended as a substitute for professional advice of any kind. Glitch.news assumes no responsibility for the use or misuse of this material. All trademarks, registered trademarks and service marks mentioned on this site are the property of their respective owners.
