MIT study links data centre power use to 650 premature US deaths in 2023

Nov 19, 2026
Pollution from the power plants that serve data centres hit Midwestern states hardest, even though most demand was in Virginia and Texas.

Air pollution from the electricity used by US data centres caused an estimated 650 premature deaths in 2023, and many of them occurred far from the facilities themselves, according to a study by researchers at MIT published in Environmental Research Letters.

The team built a model that traces how data centre electricity demand moves through the interconnected grid, identifies which power plants raise output in response and simulates how the resulting pollution forms and travels. Of the 650 estimated deaths, 513 were linked to fine particulate matter and 137 to smog, about 5.3% of all US deaths related to electricity generation.

Premature death rates were highest in Iowa, Nevada, South Dakota and Kansas, even though most data centre demand was concentrated in northern Virginia and Texas, where rates were only about half those of most Midwestern states. Generators that respond to demand in Virginia are likely in the Ohio Valley, which has a higher share of older coal and gas plants, the researchers say. "Even if you don't think you are affected because there is not a data center near you, think again," said senior author Noelle Selin, according to MIT News.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.

Also, consider following us on social media:

Free newsletter

Silicon Brief

Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more
Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.