Open data

Download, cite, build on it

Every dataset GridWatch publishes — moratoriums, projects, facilities, state profiles, and the headline archive — is free to download, licensed CC BY 4.0, and documented with a schema so you can actually use it. Researchers cite what they can query; journalists cite what they can download.

6datasets
573total records
CC BY 4.0license
Dailyrebuild cadence

Moratorium & community action tracker

307 data center moratoriums, bans, and community actions across 37 states. Each row carries a primary source, verification date, and derived expiry status — so a lapsed moratorium is never cited as current.

Download: JSON · CSV · JSON 194 KB · CSV 103 KB

307 records · CC BY 4.0 · source page

Schema — 15 fields
ColumnDescription
localityTown, city or county taking the action; 'X (statewide)' for state-level
stateUSPS two-letter code
levelLocal or State
statusAs recorded: Enacted, Proposed, Rejected, Vetoed or Rescinded
effective_statusStatus today. Equals `status`, except an Enacted row past its `expires` date reads Expired. Use this one
expiredtrue when a documented term has run out
days_leftDays until `expires`; negative once past, null when no end date is recorded
whenHuman-readable date of the action, as reported
expiresISO date the term lapses, or null when permanent, condition-based, or not documented
noteScope, threshold, vote count and other detail
sourceURL this row was read from, or null if unverified
as_ofDate the source was read, or null
verifiedtrue when a source is recorded. false means nobody has checked this row — treat it as a lead
latLatitude, null for state-level rows
lonLongitude, null for state-level rows
Cite this dataset

AI GridWatch. (2026). U.S. data center moratorium & community action tracker [Data set]. CC BY 4.0. https://aigridwatch.com/open-data

Project intelligence tracker

131 identified data center proposals tracked from rumor to hearing to decision. Each project carries a dated intelligence log so you can see what changed and when.

Download: JSON · CSV · JSON 732 KB · CSV 66 KB

131 records · CC BY 4.0 · source page

Schema — 23 fields
ColumnDescription
idStable slug, unique per project.
nameProject or code name.
operatorOperator/developer, or blank if unknown.
ownerOwner/financier, or blank.
tenantNamed end tenant, or blank.
filing_llcShell LLC on the filing (ties to the campus registry), or blank.
localityTown or county the project sits in.
stateTwo-letter state abbreviation.
latApproximate locality latitude, or blank.
lonApproximate locality longitude, or blank.
size_mwPlanned IT/critical load in MW, or blank.
acresSite acreage, or blank.
announcedISO date first publicly proposed, or blank.
rezoning_filedISO date the rezoning/permit application was filed, or blank.
hearing_dateISO date of the next/decisive public hearing, or blank.
decided_dateISO date a terminal decision was reached, or blank.
outcomeapproved | denied | withdrawn | blank (blank = still live).
stageDERIVED status shown on the page (Hearing scheduled, Awaiting decision, In review, Proposed, Rumored, Approved, Denied, Withdrawn).
days_to_hearingDERIVED days until the hearing (negative once past), or blank.
noteOne-line plain-language status.
sourcePrimary URL backing the current status.
as_ofISO date the source was read, or blank.
verifiedtrue if a source is on record.
Cite this dataset

AI GridWatch. (2026). U.S. data center project tracker [Data set]. CC BY 4.0. https://aigridwatch.com/open-data

Facility registry

84 tracked data center campuses with operator, owner, tenant, and the shell LLC on local filings — so a resident can see who is really behind a proposal.

Download: JSON · CSV · JSON 30 KB · CSV 13 KB

84 records · CC BY 4.0 · source page

Schema — 10 fields
ColumnDescription
operatorPrimary operator or developer
ownerOwner or financier, if different from operator
tenantNamed end tenant, if known
locationCity, town or locality
stateUSPS two-letter code
latLatitude
lonLongitude
filing_llcShell LLC on local filings, if known
attributionHow the operator-to-LLC link was established
srcSource URL or citation
Cite this dataset

AI GridWatch. (2026). U.S. data center facility registry [Data set]. CC BY 4.0. https://aigridwatch.com/open-data

State profiles

All 50 states and D.C. — facility count, power draw, residential rate, grid carbon intensity, and water stress. The numbers a resident cites at a hearing.

Download: JSON · CSV · JSON 14 KB · CSV 4 KB

50 states & D.C. · CC BY 4.0 · source page

Schema — 8 fields
ColumnDescription
stateFull state name
abbrevUSPS two-letter code
dc_countNumber of tracked data center facilities
twh_yearEstimated annual data center electricity consumption (TWh)
major_hubsKey metro areas and operators
rate_centsResidential electricity rate (cents/kWh)
grid_carbonGrid carbon intensity (gCO2/kWh)
water_stressWater stress level (low/medium/high)
Cite this dataset

AI GridWatch. (2026). U.S. data center state profiles [Data set]. CC BY 4.0. https://aigridwatch.com/open-data

Story tracker archive

Every community-impact headline GridWatch has archived, grouped by the town or county it's about. A running record so patterns spread across months of coverage don't disappear when the feed scrolls past.

Download: JSON · JSON 1.2 MB

CC BY 4.0 · source page

Cite this dataset

AI GridWatch. (2026). Data center community-impact story archive [Data set]. CC BY 4.0. https://aigridwatch.com/open-data

Moratorium deadline alerts

Moratoriums with documented end dates approaching or recently lapsed. Also available as RSS — subscribe once, get notified before a pause expires in your area.

Download: JSON · RSS · JSON 16 KB

CC BY 4.0 · source page

Cite this dataset

AI GridWatch. (2026). Data center moratorium deadline alerts [Data set]. CC BY 4.0. https://aigridwatch.com/open-data

License

All datasets are published under the Creative Commons Attribution 4.0 International license. You may share, adapt, and build on the data for any purpose, including commercial, as long as you credit AI GridWatch and link back. Suggested attribution:

Data from AI GridWatch, licensed CC BY 4.0.

Freshness

Every file is regenerated on each site build (currently daily). The generated field in each JSON envelope is the build date. Per-row as_of fields tell you when each individual record was last verified against its source — a dataset can be freshly generated while individual rows are months old.

API access

There is no API yet — the JSON downloads are the interface. If you are building something that needs programmatic access or real-time updates, get in touch and we will figure it out.