Sources & Methodology · Adopted November 2026

How we decide what counts as a fact

These standards apply to every figure on this site, on every side of every issue. They come first on this page because they matter more than any single fact.

Our seven standards

  1. Original sources first

    Every figure links to the original report, dataset, filing or study — not to an article about it — whenever the original is publicly available. One exception: to avoid empty fields, the Data Center Tracker may show a figure from news or industry reporting while we look for the official record. Those figures are shown in italics, each linked to the report, and are subject to revision until confirmed. Aggregator and AI-written sites are never used.

  2. Every source labeled by type

    Each citation carries a label showing who produced the information, so readers can judge it for themselves. The six types are defined below.

  3. Equal treatment for opposing views

    On every issue we present the strongest sourced evidence that supporters and skeptics each point to — with the same space, the same visual styling and the same sourcing standard. Neither view gets the last word by default.

  4. Facts kept separate from opinion

    A fact is a measurement, count, official record or study finding that can be checked against its source. Opinion pieces are clearly labeled as such, paired with an opposing view, and live in their own section.

  5. Every figure dated

    Each source shows when it was published, and each page shows when it was last reviewed. Data centers change quickly; an old number is labeled as old.

  6. Uncertainty stated openly

    Where the evidence is incomplete, estimated or disputed, we say so in a “Still uncertain” note rather than presenting a single number as settled.

  7. Corrections made in public

    When we get something wrong, we fix it and record the change, the date and the reason in the corrections log below. We do not silently edit published facts.

The six source types

Labels describe who produced the information, not whether it is right. Every type can be accurate; every type can have interests. Knowing the source helps readers weigh the evidence.

Sources cited on this site, by type

Counted automatically from the list of all sources at the bottom of this page.

Government

Public agencies, legislative auditors, national laboratories, and local governments. Examples: Lawrence Berkeley National Laboratory, Virginia’s Joint Legislative Audit and Review Commission (JLARC), county budget offices. Note: a local government can also have an interest in promoting development.

Academic

Universities and independent, nonprofit research institutes. Examples: Carnegie Mellon University, Virginia Commonwealth University, Epoch AI.

Industry

Companies, trade associations and industry-funded research. Examples: a company’s own sustainability report, a trade group’s survey. Often the only source for site-specific data.

Advocacy

Organizations that campaign for or against a position — environmental groups, taxpayer groups, residents’ associations, development coalitions. Used for firsthand accounts and data not available elsewhere.

News

Reporting by news organizations and trade publications. Used when the reporting itself is the original source — for example, interviews or records a newsroom obtained.

Community

Open data built by volunteers, such as OpenStreetMap. Used for the locations of existing facilities, where no single official list exists. Coverage can be incomplete.

How the Data Center Tracker is built

The tracker combines several layers of information. Every row shows which layer it comes from with a dot rating, and every row links to its source.

  1. Fresh public data every night. Each night an automated process downloads the latest versions of two public, openly licensed databases:
    • Academic Epoch AI’s data center dataset — the largest AI computing campuses, with status and power estimates drawn from satellite imagery, permits and company disclosures. Power figures are estimates, typically within a factor of 1.4 of the true value.
    • Community OpenStreetMap — every U.S. building or campus that volunteers have mapped as a data center. This is the same source the U.S. Department of Energy’s Pacific Northwest National Laboratory used for its national data center atlas. It is the most complete free list available, but it is not verified and coverage is uneven.
  2. Proposed projects from the news. When a verified story on our News page names a proposed, under-construction or cancelled project that isn’t already listed, it is added with the figures that story reports (power, acres, building size, incentives) and linked to the story. These rows are rated ●●○ when power is reported and ●○○ otherwise, and their map pins sit at the nearest town until an exact site is confirmed.
  3. Sorted into the same fields for every site. Each record is organized into name, owner or operator, status, location, power, size and tax incentives. Counties are looked up from each site’s coordinates using the U.S. Census Bureau’s public geocoder. Building sizes from map data are footprints traced from aerial imagery, not total floor space.
  4. Duplicates removed. When the same campus appears in more than one source, the more detailed record is kept.
  5. Highly detailed rows are never overwritten. Rows rated ●●● are checked by hand against original documents, including tax incentive agreements. The nightly process does not change them.
  6. Safeguards. If a source fails to download, the previous night’s data is kept. If the table would suddenly shrink by more than 30 percent, nothing is published until the cause is checked.
  7. Searched and sorted without favor. Every row is shown by the same rules. The default order is by status, then by how much is verified, then by size. Readers can search and filter by status, state and detail level. No project is promoted, hidden or ranked by anyone’s preference, and no company or group can pay to change how it appears.

Reported figures and images. Where a row’s fields are empty, we fill them, largest projects first, from news and industry reports. Each one is linked and shown in italics until an official record confirms it; when reports differ, all are shown. Each row’s image is a credited photo (company-provided, public-domain or freely licensed) where we have found one; otherwise it is an aerial view from the U.S. Geological Survey’s National Map and the U.S. Department of Agriculture’s National Agriculture Imagery Program (public domain), centered on the mapped location and possibly 1–3 years old.

The tracker’s data file is public. Location data from OpenStreetMap is available under the Open Database License; Epoch AI data is used under Creative Commons Attribution.

How often our data is updated

Some sources change daily; others are one-time studies. This table shows how often each source publishes new information and how often we check it.

DataSourceHow often the source publishesHow often we check
Tracker — locations of all facilitiesCommunity OpenStreetMapContinuously, as volunteers add sitesEvery 24 hours, automatically
Tracker — power and status of major campusesAcademic Epoch AIEvery few weeksEvery 24 hours, automatically
Document LibraryGovernment records, research, SEC filings, industry and advocacy documentsVaries — company filings yearly or at an IPO; reports irregularlyWeekly, automatically: newest SEC filings found, links checked, dated backup copies of public records saved. New documents added by hand.
Tracker — highly detailed projectsGovernment records, company documentsWhen a project or agreement changesUpdated by hand; every change reviewed before posting
Water use and electricity shareGovernment Lawrence Berkeley National LaboratoryIrregularly — major reports in 2016 and 2024, update in 2025Monthly
Electricity prices and demandGovernment Berkeley Lab & BrattleIrregularlyMonthly
Projected electricity costsAcademic Carnegie Mellon & NC StateOne-time study (2025)Monthly
Jobs and noise in VirginiaGovernment Virginia JLARC, Report 598One-time study (2024)Monthly
State tax incentivesGovernment Virginia JLARC incentive reportsYearly report seriesMonthly
Local tax revenueGovernment Loudoun CountyYearly, with each budgetMonthly
Company water useIndustry Google Environmental ReportYearlyMonthly
Closed-loop campus water useGovernment City of Racine recordsOne-time release (2025)Monthly
Backup generator emissionsAcademic Virginia Commonwealth UniversityOne-time study (2026)Monthly
Noise limits and reference levelsGovernment CDC / NIOSH · News Kent County coverageOnly when guidance or ordinances changeMonthly
Land and campus sizeNews Farm Policy News · Mission CriticalOne-time articles (2026, 2022)Monthly

When a monthly check finds a new edition, the proposed changes are reviewed before they appear on the site, and the change is recorded in the corrections log if it alters a published figure.

How we choose which facts to show

  1. Preference order. When several sources cover the same question, we prefer, in order: government audits and national laboratory data; academic studies; official records obtained by the public; company and trade disclosures; advocacy and news accounts.
  2. Strongest case for each view. For each issue we look for the best-supported evidence each view relies on — not the weakest version of either argument.
  3. Same source for opposing views when possible. When one study supports both perspectives (as with the Virginia noise and air quality findings), we use it for both.
  4. Balance check. Before publishing, we compare the source types behind each view on every issue. If one view relies mostly on industry or advocacy sources while the other relies on government or academic sources, we look for stronger evidence or note the imbalance.
  5. Plain language. Figures are rounded and explained in everyday terms, with the exact figure available at the source.

Waiting on original sources

These figures are held back or marked pending until we can link the original document.

Corrections log

No corrections since launch. Each future correction will be listed here with the date, what changed and why.

Independence

We take no funding from data center developers or opposition groups, and accept no advertising from any company connected to data centers.

All sources cited

Every source currently used on the site, grouped by type.

Government

Academic

Industry

Community

  • OpenStreetMap contributors. OpenStreetMap, queried nightly for U.S. features tagged as data centers. Licensed under the Open Database License (ODbL); the tracker’s location data is available under the same license. Method follows the U.S. Department of Energy’s IM3 Open Source Data Center Atlas (Pacific Northwest National Laboratory, 2025). County names from the U.S. Census Bureau geocoder. Used for: Tracker — locations and building footprints of existing facilities.

News