Escape Meridian

MIT Technology Review · Dec 17, 2024

MIT Technology Review reports rising data-center emissions amid AI’s energy demands

Markdown · JSON · Original source

On December 17, 2024, MIT Technology Review reports that data-center emissions are rising as AI companies build larger, more energy-intensive models. Research led by a team at the Harvard T.H. Chan School of Public Health found that US data-center carbon emissions have tripled since 2018, placing them slightly below domestic commercial airlines as a pollution source. The publication says this leaves AI companies caught between sustainability goals and competition to build bigger models; it expects energy-intensive models, including video generators like ’s Sora, to push emissions higher. The publication describes nuclear energy as one option companies are pursuing, while noting that nuclear plants take ages to come online. In the meantime, it says AI companies will continue to use power from sources that are far from renewable. The researchers found that most US data centers are located where electricity sources are dirtier than the national average, with the carbon intensity of energy use higher in coal-producing regions such as Virginia. The scale of future emissions, and how quickly lower-carbon power can become available, remain unknown.

Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.

MIT Technology ReviewArticle · technologyreview.comOpen original source

Original source

MIT Technology Review · Article

Source details

Original source date: Dec 17, 2024 (2024-12-17T10:00:00Z).

Date basis: Date from the publisher’s own records, through its API.

Added to Escape Meridian: 2026-11-16T05:48:24.759Z. First observed: 2026-11-02T08:33:47.515Z. Retrieved: 2026-11-02T08:35:20.758Z.

Referenced by

Recorded citations in public readings. A reference alone does not establish agreement or influence.

No public references recorded here yet.

Reference data (JSON)

Limits

Text as retrieved on November 2, 2026; originally published December 17, 2024.

Open question

How much will AI-related data-center emissions rise if energy-intensive models keep growing while nuclear plants take to come online?

Source-stated

A team at the Harvard T.H. Chan School of Public Health found that US data-center carbon emissions have tripled since 2018.

Since 2018, carbon emissions from data centers in the US have tripled, according to new research led by a team at the Harvard T.H. Chan School of Public Health.

Source-stated

The publication says energy-intensive new AI models, including video generators like OpenAI’s Sora, will send emissions higher.

The trend toward ever more energy-intensive new AI models, including video generators like OpenAI’s Sora, will only send those numbers higher.

Source-stated

The researchers found that 95% of US data centers are built in places with electricity sources dirtier than the national average.

The researchers found that 95% of data centers in the US are built in places with sources of electricity that are dirtier than the national average.

Source-stated

The publication says the carbon intensity of energy used by data centers in coal-producing regions such as Virginia is higher than the national average.

Since so many data centers are located in coal-producing regions, like Virginia, the “carbon intensity” of the energy they use is 48% higher than the national average.

Source-stated

The publication says nuclear plants take ages to come online.

However, nuclear plants take ages to come online.

Source-stated

The publication says AI companies will continue using power from sources that are far from renewable.

In the meantime, AI companies will continue to use up power from their current sources, which are far from renewable.

Source and contributor context

MIT Technology Review · Publisher

The publisher is identified here. No individual author has been inferred.

Source media

Media retrieval status: unchecked.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

OpenAI

OpenAI develops AI models and products and publishes research, release announcements and evaluations.

TIME

TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.

Editorial provenance

Version 1. Publication ID: f74be96e-fd2c-1f6f-b6b4-11b386bc87c9.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-16T05:43:56.027Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge_admitted_by_model, second_opinion_categorical.

Observation: b05fc50607e0b00f20901788e0505ffbd63daab20cfbc37c6a67ae3128c72f40. Source text hash: 1ecedb8b6f46b611d5d25b73acb882215689f0f3e6ac1dc16e46be0b4aec69e8.

Revision and source-check history

Version 1 · 2026-11-16T05:48:24.759Z

MIT Technology Review reports rising data-center emissions amid AI’s energy demands

On December 17, 2024, MIT Technology Review reports that data-center emissions are rising as AI companies build larger, more energy-intensive models. Research led by a team at the Harvard T.H. Chan School of Public Health found that US data-center carbon emissions have tripled since 2018, placing them slightly below domestic commercial airlines as a pollution source. The publication says this leaves AI companies caught between sustainability goals and competition to build bigger models; it expects energy-intensive models, including video generators like ’s Sora, to push emissions higher. The publication describes nuclear energy as one option companies are pursuing, while noting that nuclear plants take ages to come online. In the meantime, it says AI companies will continue to use power from sources that are far from renewable. The researchers found that most US data centers are located where electricity sources are dirtier than the national average, with the carbon intensity of energy use higher in coal-producing regions such as Virginia. The scale of future emissions, and how quickly lower-carbon power can become available, remain unknown.

Original source

MIT Technology Review · Article

Source details

Original source date: Dec 17, 2024 (2024-12-17T10:00:00Z).

Date basis: Date from the publisher’s own records, through its API.

Added to Escape Meridian: 2026-11-16T05:48:24.759Z. First observed: 2026-11-02T08:33:47.515Z. Retrieved: 2026-11-02T08:35:20.758Z.

Limits

Text as retrieved on November 2, 2026; originally published December 17, 2024.

Open question

How much will AI-related data-center emissions rise if energy-intensive models keep growing while nuclear plants take to come online?

Source-stated

A team at the Harvard T.H. Chan School of Public Health found that US data-center carbon emissions have tripled since 2018.

Since 2018, carbon emissions from data centers in the US have tripled, according to new research led by a team at the Harvard T.H. Chan School of Public Health.

Source-stated

The publication says energy-intensive new AI models, including video generators like OpenAI’s Sora, will send emissions higher.

The trend toward ever more energy-intensive new AI models, including video generators like OpenAI’s Sora, will only send those numbers higher.

Source-stated

The researchers found that 95% of US data centers are built in places with electricity sources dirtier than the national average.

The researchers found that 95% of data centers in the US are built in places with sources of electricity that are dirtier than the national average.

Source-stated

The publication says the carbon intensity of energy used by data centers in coal-producing regions such as Virginia is higher than the national average.

Since so many data centers are located in coal-producing regions, like Virginia, the “carbon intensity” of the energy they use is 48% higher than the national average.

Source-stated

The publication says nuclear plants take ages to come online.

However, nuclear plants take ages to come online.

Source-stated

The publication says AI companies will continue using power from sources that are far from renewable.

In the meantime, AI companies will continue to use up power from their current sources, which are far from renewable.

Source and contributor context

MIT Technology Review · Publisher

The publisher is identified here. No individual author has been inferred.

Source media

Media retrieval status: unchecked.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

OpenAI

OpenAI develops AI models and products and publishes research, release announcements and evaluations.

TIME

TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.

Editorial provenance

Version 1. Publication ID: f74be96e-fd2c-1f6f-b6b4-11b386bc87c9.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-16T05:43:56.027Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge_admitted_by_model, second_opinion_categorical.

Observation: b05fc50607e0b00f20901788e0505ffbd63daab20cfbc37c6a67ae3128c72f40. Source text hash: 1ecedb8b6f46b611d5d25b73acb882215689f0f3e6ac1dc16e46be0b4aec69e8.

uncertain · 2026-11-16T05:43:56.027Z · anthropic/claude-opus-5.5

Publication: f74be96e-fd2c-1f6f-b6b4-11b386bc87c9

attribution — pass: The reading attributes the reporting to MIT Technology Review. It attributes the emissions findings to research led by a Harvard T.H. Chan School of Public Health team, as the source does. The page is one newsletter piece, not a listing; the related-story links below it are page furniture.

This story originally appeared in The Algorithm, our weekly newsletter on AI.

summary — uncertain: The reading is mostly faithful and readable, but it subtly reshapes one point. The source says data centers overall draw energy with 48% higher carbon intensity than the national average because many sit in coal-producing regions. The reading instead says carbon intensity is higher in coal regions such as Virginia. The term "carbon intensity" (emissions per unit of energy) is central, yet it is not explained and no definition is supplied. Smallest repair: state that the energy data centers use is 48% more carbon-intensive than average because many sit in coal regions, and add a tooltip for the term.

Since so many data centers are located in coal-producing regions, like Virginia, the “carbon intensity” of the energy they use is 48% higher than the national average.

date — pass: The reading's date, December 17, 2024, matches the UTC day of originalPublished and the URL path. Some later page items, such as a Deep Dive story referring to ten years after AlphaGo, postdate the piece; the reading neither uses nor relies on them. The forecast about Sora-type models pushing emissions higher is kept as the publication's expectation. This is a publisher page, not an archive copy, so no archive disclosure is needed.

support — uncertain: Claims 1, 2, 3, 5 and 6 match their excerpts. Claim 4 shifts scope: the source applies the higher carbon intensity to data centers as a whole, explained by their location in coal regions. Claim 4 recasts this as data centers in coal-producing regions having higher intensity, which the excerpt does not state as such. This is a scope change, not a clear contradiction. The closing unknowns and the question follow from the source without adding unsupported benefits or costs.

the “carbon intensity” of the energy they use is 48% higher than the national average

media — pass: No media is selected, and no verified candidates were supplied, so there is no identity or relevance issue.

Challenge or correct this reading

Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.