THE LOADSTAR NUMBER · FREE TO CITE
reading the generation queues live from official ISO data…·
How much of the power supply pipeline behind the AI-era buildout shows a signed agreement — measured live across official U.S. ISO generation queues.
Measured live from official U.S. ISO interconnection queues — PJM · ERCOT · CAISO · MISO · SPP · ISO-NE · NYISO. Every figure links to its filing and is recomputable.
Citation copied — paste it anywhere.
⚡ The energy map
Where AI power is forming — by region.
AI's #1 bottleneck is power. The Energy Map ranks U.S. regions by how much low-carbon generation is forming in each region's interconnection queue — the power supply chasing the AI buildout, straight from the grid operators' own data.
Descriptive of today — the queue forming now, not a forecast.
Who pays
The AI data-center boom is already on your power bill.
Utilities and capacity markets price forecast data-center demand into bills now, and across the largest U.S. grid that added billions in a single year. That is the real, rising cost of the AI boom — and the part no one can yet check is how much of the demand behind it is committed, and how much is speculative.
PJM · 13 states + DC
$9.3 billion in one year
the data-center-attributable share of the rise in PJM consumers' capacity cost for one delivery year (63% of the total increase), per PJM's own market monitor.
Source: Monitoring Analytics ↗Virginia
+$11.24 a month
the 2026 residential increase Virginia's regulator approved in Dominion's rate review — and state analysts project a further $14–$37 a month by 2040 driven by data-center growth. From 2027, a new data-center rate class makes large loads cover most of their own costs.
Source: SCC · JLARC ↗The blind spot
no public field says which is real
Generation queues generally publish agreement status — in our reading, a machine-readable field exists in most of the seven — while large-load requests disclose nothing, so from the public record alone, no single field separates a real data center from a speculative one.
The dollar figures are on the public record, reported by regulators and grid monitors, not measured by us. What we add is the missing read: how much of the demand driving those bills is corroborated today, region by region, with the receipt attached.
See where the buildout is actually real →On the record
9 predictions, sealed and dated —
public since June 18.
We don't just publish a number. Every regional call is timestamped and hash-sealed the moment we make it, chained so it can't be edited after the fact, and resolved in public — hitsand misses. The commitment is already on the record; the verdicts landDec 15, 2026.
Sealed calls · sample
hash-sealed · June 18, 2026
- Central Indiana932026-06-1893b5c57f…
- Silicon Valley762026-06-1874090dab…
- Phoenix Metro452026-06-183347998d…
Score 0–100: real ·unsure ·phantom. We seal the low ones too.
Why we exist
AI is being built faster than the power to run it.
Every week brings another gigawatt-scale data-center announcement. But a press release isn't a power plant, and the grid can't tell the difference. Utilities commit billions in upgrades, regulators approve rate hikes, and the public foots the bill — all against a queue of power requests nobody verifies.
We built LOADSTAR to measure that gap: how much of the AI power pipeline is real, region by region, straight from official filings. Every number is sealed and dated, so it can't move after the fact. We take no side and sell no project. We publish what's forming, and what's phantom.
Others sell the answer privately to one developer. We publish the number for everyone: free to cite, sourced, and neutral.
How we stay neutral →A living index
The index, building itself. Live.
LOADSTAR is a living index. The number reads live from its source, the machine runs in public around the clock, and corrections are published in the open, dated, in a ledger anyone can recompute. The numbers here are live and verifiable. Nothing is a mockup.
Who needs to know if it's real or phantom
Utilities
Before investing capex in infrastructure, it's worth knowing whether the data center will actually get built.
Investors and funds
Before allocating capital to a physical asset, it's worth knowing whether the announced demand is real.
Regulators
Before approving load requests, it's worth knowing which ones are speculative.
How we get to a number
Collection
Tens of thousands of documents a day, from public sources: SAM.gov, SEC EDGAR, regional grid operator interconnection queues, USPTO, GDELT.
Filter
Deterministic rules (not AI) cut the volume down to a handful of genuinely relevant documents, before any analysis.
Correlation
We cross-reference independent signals: environmental permit, energy contract, hiring, federal registration. The more signals converge, the higher the score.
Proof
Every result is dated and sealed in an immutable ledger. The history is public — including when we're wrong.
We don't ask you to trust us.
We ask you to check.
We don't say whether it's real — we say how much we were able to corroborate. Every headline number links to its source. Every score is sealed by hash, dated, recomputable by anyone.
Absence of signal isn't proof of phantom. It's a coverage gap — and we say which one.
The Reality Report
A dated reality check on the AI buildout — every week.
Region-by-region scores, the honest read behind the numbers, and the receipts. Free, and cross-posted in full on Substack.
Start with the sample report
No commitment. A real report, on a real project, with the methodology exposed start to finish.