AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: Compute Campuses, Utilities, And The Costs OpenAI Hasn’t Settled on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get audio and creator gear delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

OpenAI confirmed it is using the term “compute campuses” for its AI infrastructure, while a broader post-election public relations campaign has been reported but not confirmed by the company. The costs of powering and connecting large data centers are already visible in PJM’s capacity market, but how much households ultimately pay—and how enforceable industry promises are—remains unclear.

OpenAI has confirmed it is using “compute campuses” to describe the large facilities it is building for artificial intelligence, as questions persist over who will pay for the electricity, grid upgrades and water those projects require. Axios reported that the company is preparing a broader public relations effort after the U.S. midterm elections, but OpenAI has confirmed the terminology—not that campaign.

OpenAI told Business Insider the new phrase “better describes the scale and role of the infrastructure we’re building for AI.” The company also said the facilities involve “long-term investments in computing, energy infrastructure and skilled work,” and that host communities should help shape projects and share in their opportunities. The term’s confirmation does not settle the practical question behind the debate: how infrastructure costs are allocated among developers, utilities and other customers.

The scale of the issue is apparent in the PJM electricity market, which serves more than 67 million people across 13 states and Washington, D.C. Its July 2026 capacity auction for the 2028–29 delivery year cleared at the regulator-approved cap of $325 per megawatt-day. The auction was 6,831 megawatts short of the capacity PJM sought. The market’s independent monitor attributed $6.3 billion of a $16.4 billion total to data centers, according to the source material.

Those figures show data centers are a substantial part of the capacity-market bill, but they do not establish that data centers alone caused household electricity-price increases. The source material says residential prices rose 11.3% in Virginia and 10.4% in Ohio over a year, while the International Energy Agency found no simple state-level relationship between data centers and household prices. The Energy Information Administration attributed a 41% increase in PJM wholesale prices to weather, not data centers.

At a glance
reportWhen: Term confirmed November 2026; cost and…
The developmentOpenAI confirmed the term “compute campuses” as scrutiny grows over the electricity, grid and water costs of large AI infrastructure projects.
Video: OpenAI Wants to Call Data Centers “Compute Campuses”. Who Pays for the Power and Water?. Narration: AI voice clone of Thorsten Meyer; images AI-generated.

Who Pays for AI Power

The costs matter beyond the companies building AI infrastructure because power plants, transmission lines and local distribution systems serve both new industrial demand and existing customers. If project-specific costs are not fully recovered from developers, other utility customers could help pay; if costs are assigned too broadly, developers could also face charges that do not reflect their actual use. The IEA says the distribution of risk depends on local rules and rate design.

OpenAI has pledged to “pay our own way on energy,” and seven technology companies signed a White House pledge covering new power generation and delivery infrastructure. Those commitments could help address community concerns, but the available fact sheet does not specify penalties, an oversight body, monitoring or required public reporting. That leaves a gap between stated commitments and independently verifiable outcomes.

Water is another local concern, especially where facilities are large or communities already face resource constraints. At OpenAI’s planned site in Pike County, Ohio, the company expects water use comparable to an office building, but says it will publish projections once the design is final. Until those figures and the project’s operating plans are available, residents cannot assess that expectation against a completed estimate.

Products Worth Considering

Bose SoundLink Plus Portable Bluetooth Speaker, Wireless Outdoor Speaker, Up to 20 Hours Battery Life, Waterproof and Dustproof, Black
EcoFlow Delta 3 Ultra 3072Wh Portable Power Station – 3600W AC Output

EcoFlow Delta 3 Ultra 3072Wh Portable Power Station – 3600W AC Output

As an affiliate, we earn on qualifying purchases.

From Naming to Infrastructure Costs

The terminology emerged amid growing public concern about data centers. Axios reported on November 1 that OpenAI was preparing a post-midterm “PR blitz” and considering “AI compute campuses” as a replacement phrase. That campaign account is secondhand here: the source material says it relies on TechCrunch’s summary of Axios, and OpenAI has not confirmed that a campaign has launched. The company did confirm the term to Business Insider.

Polling cited in the source material indicates why the issue has drawn attention. Pew Research Center surveyed 10,548 U.S. adults from July 20 to August 9, 2026. Half said data centers were mostly bad for home energy costs, compared with 38% in January; 54% said they were mostly bad for the environment, up from 39%. Sixty percent said they would be not too or not at all comfortable with a new data center in their area.

OpenAI’s January 2026 “Stargate Community” plan said the company would pay its own way on energy so its operations would not raise local electricity prices. On March 4, the White House announced the Ratepayer Protection Pledge, signed by Amazon, Google, Meta, Microsoft, OpenAI, Oracle and xAI. It calls for companies to bring or buy new generation, cover required power-delivery upgrades and pay negotiated rates even if they do not use the electricity. The fact sheet describes the pledge as a “first step.”

Promises Without Public Verification

It is not clear whether OpenAI will formally launch a post-election communications campaign or how extensively it will use the new term. The company has confirmed the phrase, while the reported campaign plan comes through Axios reporting summarized by TechCrunch. The source material says no official OpenAI publication launching the terminology had appeared as of November 3.

The enforceability of the industry’s energy commitments is also unresolved. The White House pledge’s fact sheet names no penalty, enforcement body, monitoring process or reporting requirement, and the IEA says examples of U.S. states legislating full cost recovery from data-center operators remain few. The available information does not establish how each signatory will meet the pledge or how the public can check its performance.

For Pike County, Ohio, OpenAI has not yet published water-use projections for its planned site; the company says it will do so once the design is final. The source material also says OpenAI owns no data centers and relies on partners to carry project debt. A reported reduction in OpenAI’s funding target is not confirmed by the company, so its financing implications remain uncertain.

Final Designs and Cost Rules

The next concrete information for the Pike County project is expected when OpenAI finalizes its design and publishes water-use projections. Those figures should help clarify the company’s estimate, although the source material does not give a publication date or describe a completed independent review.

For electricity costs, the key developments to watch are how utilities and regulators assign expenses for new generation and grid connections, and whether states adopt rules requiring data-center operators to cover their full costs. Public reporting on implementation of the White House pledge would also show whether its commitments translate into measurable protections. Until then, the scale of infrastructure spending is clearer than the final division of its costs.

Key Questions

Has OpenAI officially changed the name of its data centers?

OpenAI confirmed to Business Insider that it is using “compute campuses” to describe its AI infrastructure. The source material says it had not found an official OpenAI publication launching the term as of November 3.

Has OpenAI confirmed a post-election public relations campaign?

No. Axios reported that OpenAI was preparing a campaign after the midterm elections, according to a report summarized by TechCrunch. OpenAI confirmed the terminology but, in the source material, did not confirm that the wider campaign had launched.

Are data centers responsible for higher household electricity bills?

The relationship is disputed and varies by location. The source material reports rising residential prices in Virginia and Ohio, but says the IEA found no simple state-level relationship and the EIA attributed a 41% PJM wholesale-price increase to weather.

What does the White House Ratepayer Protection Pledge require?

The pledge calls on seven companies, including OpenAI, to bring or buy new power generation, pay for required delivery infrastructure upgrades and pay negotiated rates whether or not they use the electricity. The fact sheet does not specify penalties or a monitoring and reporting system.

When will OpenAI publish water estimates for its Ohio project?

OpenAI says it expects to publish water-use projections for its planned Pike County site once the design is final. The source material does not provide a date for that design or for publication.

Source: ThorstenMeyerAI.com

You May Also Like

2026’S Top AI Student Planners For Better Time Management And Results

Discover the leading AI-powered student planners of 2026, designed to improve time management and academic results for students at all levels.

Jumpstart Your Applied Research With Ilya’s 30 Recommended ML Papers

A new curated list of 30 beginner-friendly ML papers by Ilya aims to help R&D leaders quickly identify research with commercial potential, streamlining innovation.