Legal
Data Disclaimer
Last updated July 25, 2026. What our numbers are, what they are not, and what you should do before relying on them.
Every figure published by DC Atlas is an estimate. Capacity, power, land, floor area, cost, ownership, tenancy, project timing and market totals are compiled and derived from publicly available information. They are not audited, they are not warranted, and they will sometimes be wrong.
We work hard to be accurate, and we publish how we do it so you can judge the work rather than take it on faith. That effort is not a guarantee. This page sets out the limits of what we are telling you, and it applies to everything we publish: the website, the application, exports, reports, the API, alerts, and anything generated by Atlas AI.
1. All values are estimates
Data centers are private infrastructure. Operators are under no obligation to publish their capacity, their utilisation, their costs or their customers, and most do not. What is public is fragmentary: a press release here, a planning application there, a utility filing, an investor deck, a job advertisement.
DC Atlas assembles those fragments into structured records. Where a figure has been published by an operator we record it and say so. Where it has not, we may derive an estimate from what is known, for example from building footprint and typical power density, and we label it as an estimate. Both kinds of figure carry the same caveat: they describe our best current reading of the public record, not a measured fact.
- Capacity figures are stated in megawatts of IT load unless a page says otherwise, and different sources measure power in different ways.
- Live capacity and full buildout are held as separate figures. An announced ambition is never presented as operational supply.
- Pipeline records describe announced, permitted or under-construction projects. Projects are delayed, resized, relocated and cancelled routinely.
- Commercial values, where shown, are derived from public filings and reporting. They are not company accounts and they are not reconciled to any audited statement.
- Aggregates such as market and operator totals are computed from our own records, so any gap or error in those records flows through to the total.
2. Where the information comes from
Our records are grounded in publicly available sources: operator websites and marketing material, press releases and investor communications, planning and permitting filings, regulatory and grid-operator disclosures, public procurement notices, and reputable industry press. We use automated and AI-assisted research to find, cross-reference and summarise that public information at scale.
What we publish is not those sources, it is our dataset. Resolving thousands of fragmentary references into one record per facility, campus, operator and market takes real work: deduplicating, reconciling sources that disagree, deriving our own estimates where nothing has been published, and maintaining all of it as it changes. That work is ours, and the dataset it produces is proprietary to DC Atlas. It does not change anything below. Our own work is still an estimate, and it still needs verifying.
That means our data inherits the limits of its sources. A source can be outdated, promotional, mistaken, or silent. Automated extraction can misread a document. Cross-checking reduces those errors; it does not eliminate them. Where sourced values are shown with a citation, the citation tells you what a document said on the date we captured it, not that the statement was true.
3. Provided as-is, with no warranty
The data is provided on an “as is” and “as available” basis. To the fullest extent permitted by law, DC Atlas makes no representation and gives no warranty, express or implied, that any information is accurate, complete, current, reliable, or fit for any particular purpose. We do not warrant that a facility exists as described, that it is owned or operated by the party we name, that its capacity is as stated, or that any planned project will be built.
We may change, correct or remove any record at any time. Historical snapshots reflect what we held at the time, not a restatement of the past.
4. You must do your own due diligence
DC Atlas is a research starting point. It is built to tell you where to look and what to ask, and it is not a substitute for verification. Before you act on anything you find here, confirm it independently with the operator, owner, utility, authority or adviser concerned, and with primary documents.
This matters most in exactly the situations our users care about: investment and acquisition decisions, site selection, capacity procurement, lending and valuation, competitive analysis, and any public statement that repeats one of our figures.
5. Nothing here is advice
Nothing published by DC Atlas constitutes investment, financial, legal, tax, engineering, real-estate or site-selection advice, and nothing is a recommendation to buy, sell or hold any security or asset, or to enter into any transaction. We are not a registered investment adviser, broker, valuer or engineer, and no fiduciary or advisory relationship arises from your use of the service.
Content generated by Atlas AI is produced by automated language models from the same underlying records. It can be incomplete or wrong in ways the records are not, and it must be checked against the underlying data and the original sources before it is relied on or repeated.
6. Forward-looking information
Pipeline records, buildout figures, project timelines and anything else describing the future are forward-looking by nature. They reflect announcements and filings made by third parties, which are themselves statements of intent rather than commitments. Actual outcomes routinely differ, often by years and by hundreds of megawatts. Do not treat a pipeline figure as committed supply.
7. Limitation of liability
To the fullest extent permitted by law, DC Atlas is not liable for any decision made, action taken, or loss suffered in reliance on the data, including any loss of profit, revenue, opportunity, goodwill or anticipated saving, and including any indirect or consequential loss. Your use of the data is at your own risk, and the limitations and exclusions in our Terms of Service apply to this page in full.
Nothing in this disclaimer excludes or limits any liability that cannot lawfully be excluded or limited, including liability for fraud or fraudulent misrepresentation.
8. Third-party names and trademarks
Company names, facility names, trademarks and logos referred to on DC Atlas belong to their respective owners and are used descriptively, to identify the subject of a record. Their appearance does not imply any affiliation with, sponsorship by, or endorsement from those owners, and it does not imply that they have reviewed or approved anything we publish.
9. Corrections and removal requests
If we have described a facility, a company or a project incorrectly, tell us and we will review it against the public record and correct it. Corrections go through the same verification process as everything else, and the previous value is retained in our history. Write to contact@dcatlas.io with the record and the correction, and a source where you have one.
If you believe a record discloses something that should not be public, say so in the same message and explain why. We assess removal requests on their merits, including any security-sensitivity a site operator raises.
10. Your acceptance
By using DC Atlas you confirm that you have read this disclaimer, that you understand our figures are estimates, and that you accept responsibility for verifying anything you rely on. If you do not accept it, do not use the service.