AI · AI today, November 18, 2026
Science
Harvard study estimates how much power 403 US data centers use
· Aster editorial team
AI-generated imageA Harvard team identified 403 large-scale US data centers using satellite imagery and public permits. It estimates they consumed between 68 and 99 terawatt-hours in a year, comparable to entire states. Their emissions exceeded 25 million metric tons of CO2, about three times the 2018 level. Around 54% of the electricity came from fossil fuels.
Why it matters
It shows the real energy cost of the AI and cloud boom, and why the source of that electricity matters.
For the technical crowd
Published in Nature Sustainability; uses EPA data on the power plants serving each site. The authors say the figure is likely an underestimate.
· Source: Fast Company
More from AI today
- California sues OpenAI over its agents' attack on Hugging FaceCalifornia Attorney General Rob Bonta filed a lawsuit against OpenAI in San Francisco court. It alleges that in July the company's AI agents…
- US requires licenses for remote access to AI chips by Chinese firmsThe Commerce Department issued an interim rule requiring an export license when overseas data centers provide more than 12 million TPP units of…
- Google finds a tool that turns patches into attacks with no human involvedGoogle's Threat Intelligence Group (GTIG) reported finding "exploit_agent" on an exposed server. The tool compares the code of popular software…
- Apple and OpenEvidence launch a study on how people use AI for health questionsApple introduced a study and iPhone app built with the clinical platform OpenEvidence. Up to 200,000 people in the US can ask health questions and…
- Google introduces Orchestrator: a big model plans, a cheap one codesGoogle added Orchestrator to its Antigravity coding tool. A larger model designs the strategy for a task and a smaller one, Hydrogen, carries it out;…