U.S. Data Centers Used Nearly 2% of Electricity in One Year
A study of 403 U.S. hyperscale data centers found they consumed nearly 2% of the nation’s electricity from May 2024 through April 2025, as demand tied to artificial intelligence expands. Most of the power came from fossil fuels, according to the research published in Nature Sustainability.
The study provides peer-reviewed estimates of data-center energy use and emissions during the AI boom, a period when researchers and policymakers have struggled to obtain information from a notoriously opaque industry. Its findings cover hyperscale facilities built to support large-scale computing, including AI workloads.
The article says the facilities’ electricity use has helped spur a natural gas boom even as technology companies fund research into emerging clean-energy options such as nuclear fusion and space-based solar power. Company figures cited in the article show carbon emissions rose year over year from 2025 to 2026 by 25% at Google and Microsoft and 16% at Amazon. Together, the three companies emitted 19 million metric tonnes of carbon dioxide equivalent over the last fiscal year, roughly a third of France’s emissions.
The rising demand has also become a political issue amid concerns about consumer energy prices. The article reports that scrutiny and policy oversight of the sector are expected to intensify after the latest U.S. elections. At the United Nations climate conference in Antalya, Türkiye, only a handful of small countries had not signed a pledge to develop and power AI infrastructure and hardware sustainably.
Despite debate over whether the AI boom is a bubble, energy and finance figures cited in the article expect power demand and investment to remain high. Bloom Energy’s KR Sridhar called the outlook “nothing but a huge opportunity.” Goldman Sachs has not changed its projection of $1.4 trillion in AI-related hardware spending by the end of 2027, while BloombergNEF data-center head Mark Daly said the midterm elections offered no concrete reason to significantly revise the firm’s views.
Comments (0)
No comments yet. Be the first!