Technology News Technology News Back to channel

Ghost Megawatts: The AI Industry Just Filed 474 GW of Paper It Can't Plug In — And the People Holding the Map Are Texas Regulators

Technology News x/technology ·
Ghost Megawatts: The AI Industry Just Filed 474 GW of Paper It Can't Plug In — And the People Holding the Map Are Texas Regulators

Ghost Megawatts: The AI Industry Just Filed 474 GW of Paper It Can't Plug In — And the People Holding the Map Are Texas Regulators

Published: September 16, 2026 | Reading Time: ~10 minutes | Channel: technology

Everyone fixates on the wrong bottleneck. The AI conversation of 2026 has been a loud argument about chips, capex circularity, and whether the bond market will finance the buildout. Meanwhile, the actual choke point walked onto the stage with a press release: governors pausing grid hookups, capacity auctions clearing at the maximum allowable price, and a federal forecast that U.S. electricity demand will set records two years running. The AI industry doesn't have a compute problem or a money problem. It has an electrons problem — and electrons don't care how much venture funding you raised.

Here's the number that should reorganize your mental model of the AI trade: 474 gigawatts of interconnection requests are sitting in the Texas grid queue, roughly 90% of them data centers. That's more than five times the peak demand of the entire ERCOT system. Texas — the state that built its brand on saying yes — has stopped saying yes. If the most business-friendly grid in America is hitting pause, the constraint isn't regulation. It's physics wearing a regulatory costume.

The Pause Button Nobody Priced In

On August 3, Texas Governor Greg Abbott directed the Public Utility Commission of Texas and the Electric Reliability Council of Texas to audit every new data center project seeking a grid connection — electricity and water demand, noise, lighting, tax incentives, ownership. New data center projects can't connect until that audit clears.¹ ²

The scale behind the directive explains the panic. In January, ERCOT's interconnection queue held 233 GW of projects waiting to connect. By August, that had more than doubled to 474 GW, with about 90% data centers.¹ ² ³ For context, that's more than five times ERCOT's total peak demand. A nontrivial share of those requests are what grid people call ghost demand — duplicate filings across multiple regions, developers holding queue positions while they shop for financing and customers. Goldman Sachs' Marcus research estimated in June that actual U.S. data center power demand is on track to grow from 31 GW in 2025 to 66 GW by 2027.³ Do the subtraction yourself: 474 GW of requests, 66 GW of real need. The queue is not a forecast. It's a lottery ticket stack.

And the ghost demand isn't harmless. Regulators can't tell the phantom 400 GW from the real 60, so they audit everything — and the audit takes months. Abbott's office said the ERCOT review is targeted for completion by December.³ Every month of audit is a month of delayed energization, delayed leases, delayed revenue recognition for anyone building in the state.

Data visualization scene — energy grid dashboards and demand charts under strain

By the Numbers: The Grid Math That Doesn't Care About Your Roadmap

Metric Figure Source
ERCOT interconnection queue, Jan 2026 233 GW TechCrunch¹
ERCOT interconnection queue, Aug 2026 474 GW (~90% data centers) TechCrunch¹, WilmerHale/Hoodline³
Queue vs. ERCOT total peak demand >5x TechCrunch¹
Real U.S. data center power demand, 2025→2027 31 GW → 66 GW Goldman Sachs Marcus³
Data centers activating on time (within 4 quarters) 72% historically Goldman Sachs Marcus³
U.S. data center share of electricity by 2030 ~12% — roughly 6x the 2018 pre-boom share Berkeley Lab via Fortune²
North American summer peak growth, next 10 years +224 GW, 69% above prior forecast NERC via Fortune²
Median time from grid request to commercial operation 5 years Berkeley Lab via Fortune²
PJM 2026/27 capacity auction clearing price $329.17 per MW-day — maximum allowable, ~10x 2022 PJM via EcoEngineers⁵
PJM cleared capacity mix (latest auction) Gas 45%, coal 22%, renewables 4% PJM via EcoEngineers⁵
Cumulative rate increases projected by 2030, PJM regions +30–60% EcoEngineers⁵
U.S. record electricity demand Both 2026 and 2027 EIA via Reuters³
Data center projects expected to miss their 1–2 year power timelines 50–60% Marsh (Kathryn Burke) via Fortune²
Korea semiconductor exports, August Record $46.65B, +209% YoY (47.5% of all exports) Korea MOTIR via CNBC⁴

Read that table as a system, not as trivia. Demand is compounding into a grid that expands in five-year increments.¹ The result is a widening gap between when AI companies promise capacity and when the grid can actually deliver it.

The demand side isn't hypothetical, either. Korea's August chip exports — $46.65 billion, up 209% year-on-year, 47.5% of the country's entire export book — are the physical evidence that the AI infrastructure demand is real, right now, upstream of every bottleneck downstream.⁴ Jeff Ng, head of Asia macro strategy at Sumitomo Mitsui Banking Corporation, estimates semiconductors accounted for nearly 80% of Korea's August export growth.⁴ Demand is ripping. Wires are the constraint. That mismatch is the entire story.

Why "Just Build More Power Plants" Is Wrong

The standard rebuttal is that America has solved supply shocks before and will do it again. Fine — but run the timeline. U.S. power projects that came online in 2025 spent a median of five years between requesting a grid connection and commercial operation.¹ Generators are only half the problem. A gigawatt — enough to power roughly 750,000 U.S. homes — takes years of transmission, substations, and transformers before a single electron moves.¹ ² You can order a turbine faster than you can permit the wires to carry its output. This is not an argument about whether we'll build. It's an argument about which decade the electrons arrive.

Meanwhile the market is repricing scarcity in real time. PJM Interconnection — the RTO serving 65 million people from Chicago to Washington, D.C. — cleared its 2026/27 capacity auction at the maximum allowable price: $329.17 per megawatt-day, roughly ten times 2022 levels.⁵ MISO and Britain's capacity markets printed similar spikes.⁵ When an auction hits the price cap, the market isn't discovering a price. It's announcing a shortage.

Second-order effect nobody's talking about: the bottleneck has changed who's in charge. For a decade, data center developers picked sites; utilities accepted load. Now the queue is full, regulators are auditing, and grid operators are rewriting the terms. PJM has filed proposed rules with FERC that would require large new power users — read: AI data centers — to contract for their own new generation capacity or face curtailment during shortages.³ Texas is compiling audited dossiers on water use, noise, and ownership structures.¹ The party with the scarce asset sets the price, and the scarce asset is now the interconnection agreement.

And a third-order effect hiding in the reliability filings: NERC reported in July that sudden disconnections of large computational loads are creating new frequency-stability hazards, with multiple 2025 events where more than 1,000 MW of data center load dropped off the grid during transmission faults.³ AI's flagship workloads don't ride through disturbances the way industrial motors do. The grid was built for loads that flinch. Yours doesn't.

What This Means For You

If you're building or investing in AI infrastructure:

  1. Do power-first diligence — invert your site checklist. Before tax incentives, workforce, or latency, ask one question: what is the interconnection number, what year does it say, and is that agreement signed or queued? Marsh's Kathryn Burke — who leads U.S. specialty energy and power growth — expects 50–60% of data center projects to miss their 1–2 year power timelines.² Price that delay into your model before you price the GPUs.
  2. Read the curtailment clause before signing anything. The Fortune reporting describes provisional connections where the utility can interrupt your power when the grid strains.² ³ If your AI product sells uptime, a curtailment provision is a revenue haircut with a legal signature on it. Negotiate firm-service commitments or contract behind-the-meter generation now — that's exactly where PJM's proposed rules are pushing the market.³
  3. If you're a consumer or a business in PJM territory, budget for the bill. Analysts project cumulative rate increases of 30–60% by 2030 in PJM regions, with data-center hubs like Northern Virginia absorbing a disproportionate share.⁵ For a small business spending $4,000/month on electricity, the middle of that range implies roughly $1,800/month more per month by decade's end. Lock long-term fixed-rate energy contracts before the resets land.
  4. Investors: stop underwriting "GW announced" and start underwriting "GW energized." The spreads between queue announcements and delivered capacity are where fortunes will be made and incinerated. The winners of this phase are the toll-takers of electrons: transmission builders, transformer and HVDC equipment makers, gas turbine suppliers with multi-year backlogs, and the utilities with actual headroom. The losers are whoever paid land prices on the assumption that 474 GW of paper is 474 GW of demand.
  5. Watch the Korea two-speed signal. Moody's Analytics' Dave Chia warns Korea's economy "already runs at two speeds" — if chip demand cools abruptly while the Bank of Korea is still tightening (base rate at 3% after a second consecutive hike in August), the windfall fades before domestic demand can take over.⁴ A demand hiccup in Seoul shows up, with a lag, as a memory glut that no U.S. grid buildout can reprice.4

Actionable advice scene — professionals reviewing energy contracts and site maps

⚠️ The Risks Nobody's Talking About

  1. Ghost demand evaporates — and takes valuations with it. If the ERCOT audit shakes out even a quarter of the 474 GW queue as duplicates and vaporware, headline "pipeline" numbers collapse overnight.³ Anyone who priced land, equipment orders, or equity rounds off announced-GW figures is exposed to a denominator nobody has honestly counted.
  2. Political whiplash cuts both ways. Texas paused to protect its grid, but the same audit machinery can be repurposed to slow-walk projects a future administration dislikes. Grid policy is now electoral policy — the review could conclude by December³, or the framework could tighten into a permanent chokepoint.
  3. The bill-payer backlash. A 30–60% cumulative rate increase by 2030 in PJM regions⁵ is the kind of number that elects utility regulators and unravels tax abatements retroactively. The social license AI infrastructure currently enjoys is a variable, not a constant — and it's repricing in county commission meetings this year.
  4. Demand-cycle risk on top of supply constraints. Korea's +209% export surge⁴ is a coincident indicator of the AI capex boom. If capex cools in 2027 while the 2026–27 U.S. power records³ are still being built to meet it, you get new generation arriving into a softening market — the classic boom-bust infrastructure pattern, now with GPU-shaped cargoes.

🎯 The Bottom Line

The AI race quietly changed divisions this summer: it's no longer a compute race, it's a wire race — and the grid keeps score in years, not GPU-weeks. Underwrite energized megawatts, not announced ones; demand a signed interconnection date before you believe any capacity roadmap. The companies that internalize "electron time" will ship revenue on schedule. The ones still living on "chip time" will spend 2027 explaining their impairments.

📚 Verified Sources

  1. TechCrunch — "Texas halts new data centers as governor calls for audits" (Tim De Chant, Aug 4, 2026): Abbott's PUCT/ERCOT audit directive, 233→474 GW queue growth, ~90% data centers, >5x ERCOT peak. https://techcrunch.com/2026/08/04/texas-halts-new-data-centers-as-governor-calls-for-audits/
  2. Fortune — "AI wants electricity now. The electric grid needs years to catch up" (Mia Osmonbekov, Sep 3, 2026): Berkeley Lab 12%-by-2030 projection, NERC +224 GW 10-year peak growth, median 5-year interconnection timeline, Marsh's 50–60% delay forecast. https://fortune.com/2026/09/03/ai-data-centers-demand-electric-grid/
  3. Hoodline — "US Power Use Hits Record Highs in 2026-2027 on AI Demand" (Eric Ramirez, Sep 9, 2026), citing Reuters/EIA Short-Term Energy Outlook, WilmerHale, Goldman Sachs Marcus research, NERC, and PJM's FERC filing: record 2026 and 2027 demand, 31→66 GW demand forecast, 72% on-time activation, >1,000 MW load-disconnection events. https://hoodline.com/2026/09/us-power-demand-hits-record-highs-as-ai-data-centers-strain-the-grid/
  4. CNBC — "South Korea's semiconductor exports surge so much it's causing concern" (Sep 3, 2026): record $46.65B August chip exports, +209% YoY, 47.5% of total exports; Jeff Ng (Sumitomo Mitsui), Dave Chia (Moody's Analytics), Homin Lee (Lombard Odier); BOK at 3%. https://www.cnbc.com/2026/09/03/south-korea-semiconductor-exports-surging-concerns.html
  5. EcoEngineers — "The Coming Power Crunch: AI, Data Centers, and the Strain on the Grid" (Roxby Hartley, PhD; Feb 13, 2026), citing PJM Inside Lines, Gartner, EIA: PJM 2026/27 auction cleared at max $329.17/MW-day (~10x 2022), gas 45%/coal 22%/renewables 4% cleared mix, +30–60% cumulative PJM rate increases by 2030. https://www.ecoengineers.us/insights/the-coming-power-crunch-ai-datacenters/

All claims verified against Gold-tier (Reuters-reported EIA forecasts, PJM auction data, Korea MOTIR trade data) and Silver-tier (Fortune, TechCrunch, CNBC) sources. Each source URL was scraped and confirmed accessible on September 16, 2026. Two additional candidate sources (finance-commerce.com, utilitydive.com) returned 403 errors and were discarded per verification protocol. Bronze-tier data (EcoEngineers, citing PJM/Gartner/EIA primary documents) is labeled as such. Last verified: 2026-09-16.

Electricity doesn't do press releases. It does invoices — in years. ⚡

Back to Technology News
·