GridWatch is built on public data. This page explains exactly where our numbers come from, how we process them, and what we do — and don’t — do with other people’s work.
The data-center map
Every facility on our map and state pages comes from OpenStreetMap, the open, community-built geographic database, queried through its Overpass API for features tagged as data centers. OpenStreetMap data is published under the Open Database License (ODbL); we use the facts — locations, operators, status — the way the open-mapping community intends. We refresh the dataset hourly and keep a versioned record of every change.
OpenStreetMap is broad but not complete, and its tags can lag reality. If you know of a facility we’re missing or showing wrong, tell us — and consider adding it to OpenStreetMap itself, where the whole community benefits.
Location: state, county & city
OpenStreetMap gives us a coordinate for each facility. To place it in a state, county, and city, we run a point-in-polygon test against official U.S. Census Bureau TIGER/Line boundary files and match the nearest populated place from the GeoNames open database. That’s why our per-state counts can differ from a raw tag: we compute location from the map, not from whatever a contributor typed.
Statewide context
Our map shows the facilities we’ve pinned; it’s a floor, not a full census. For statewide totals — operational and planned facilities and projected tax revenue — we cite the American Edge Project and TECNA’s America’s AI Surge report (December 2025), and we say so on the page. Where our pinned count is lower than the statewide estimate, we tell you that, and ask for help mapping the rest.
Public records & the Docket
Our Docket is built from public government records — meeting agendas, permit filings, zoning applications, and utility dockets — collected from municipal, county, and state sources. When we publish a record, we render the public document and link to its origin. Public records are public; publishing one is not an endorsement of its contents.
The news wire
Our National Record Index is an archive of headlines and links to news published elsewhere. We link to the original source and quote only what’s necessary for reporting and commentary. We do not store or republish full copies of other outlets’ articles. Our own articles are original analysis, written under our Editorial Standards.
What we won’t do
We don’t scrape or redistribute commercial, license-restricted databases, even when they’d be convenient. Our facility base is open data, on purpose. We’d rather have a transparent, reproducible dataset than a bigger one we can’t stand behind.
Found a mistake?
Data is never perfect, and ours is a living picture of a fast-moving buildout. If a number looks wrong, tell us — and see our Corrections Policy for how we find, fix, and disclose errors.