If every candidate pledging a data center pause won on 3 November, the states where they are running would hold roughly 30 GW of capacity planned for 2028, according to a 1 November analysis by SemiAnalysis based on its Datacenter Industry Model. If the candidates who back conditions instead win, it puts the exposure at 5.7 GW.
SemiAnalysis models 62.2 GW of US net additions in 2028, most of it not yet under construction. By our arithmetic, the 30 GW equals about half of that, and the gap between the two outcomes is about 24 GW.
Business Pill · Lead Time Explained
A short explainer on lead times. It teaches the general idea only and says nothing about these elections.
The key insight: As we read SemiAnalysis’s analysis, Tuesday’s vote matters less for the 2027 data center build, which it says is mostly under way, than for 2028, where most capacity has not started and still needs state approvals.
SemiAnalysis wrote that “Roughly three-quarters of the 33.8 GW due across our top 20 states is already under construction” for 2027. It added: “The fight is over 2028. Our model has 62.2 GW of US net additions that year, most of which hasn’t started construction.”
It tracks the 13 states with the largest 2028 additions, setting aside Virginia and Mississippi, which hold no state elections this November. It says they “account for 43.5 GW, or 70% of what we expect nationally, and 10 of the 13 elect a governor”.
On the candidates, it says “nine of the 20 nominees want a pause on new construction and the other 11 want conditions, and the split is about even by party”.
SemiAnalysis Datacenter Industry Model (1 November 2026 analysis, free preview): 33.8 GW due in 2027 across its top 20 states, roughly three-quarters under construction; 62.2 GW of US net additions modelled for 2028, mostly not started; roughly 30 GW of planned 2028 capacity in states where pause candidates are running, against 5.7 GW if conditions-only candidates win.
Texas Is the Biggest Single Risk
SemiAnalysis wrote that “The biggest risk to our buildout forecast right now is a prolonged permitting freeze at the TCEQ. Texas alone adds close to 30% of US GW in 2028.”
It says Texas froze Texas Commission on Environmental Quality permits on 21 September, on top of the grid interconnection-queue audit begun on 3 August, and that “ERCOT is targeting Dec. 10 to file its report, which does not by itself reopen permitting”.
By our arithmetic, close to 30% of 62.2 GW is roughly 18.7 GW. Our earlier arithmetic put Musk’s 1.2 million Blackwells at about 2.2 GW, so the Texas share alone is about eight and a half builds of that size.
On the governor’s race, SemiAnalysis says Gov. Greg Abbott (R) could reopen permitting once the grid and water reviews finish, while his challenger Gina Hinojosa (D) “would impose a moratorium until legislation is passed that gives communities a voice and protects household utility bills, as well as ends the tax incentives for datacenters”.
It also says Oregon “expanded its earlier state-land freeze to air and water-discharge permits on Oct. 9, through July 1, 2027”.
What Candidates in Both Parties Are Asking For
SemiAnalysis lists three positions many candidates are campaigning on: “Make datacenters pay their own grid costs (15 candidates)”, “End the tax breaks (12 candidates)” and “Limit or report water use (11 candidates)”.
It quotes Rex Meyer (R), running for the Linn County Board of Supervisors in Iowa: “The public gets to know about all of this last”. Josh Tolbert (R), running for the Georgia Public Service Commission, said: “If a datacenter causes a cost, that datacenter should pay it. Existing customers’ bills should not go up because of datacenters. It is that simple”.
Pam Marsh (D), an Oregon state representative, said voter anxiety over AI is part of it: “We are starting to hear stories about people losing their jobs to AI”.
Separately, The Information published an analysis, summarised by Techmeme, that “candidates backing a data center moratorium could win up to 15 of 36 US governors’ races on Tuesday, including in Texas, Ohio, Michigan, and New York”.
The Business Lever: Approvals, Not Money
SemiAnalysis’s distinction is between candidates who would change the price of building and those who would stop approvals. As we read it, conditions such as paying for grid upgrades raise costs for developers, while a pause blocks any project that still needs a permit the state has stopped issuing.
As we read it, SemiAnalysis separates two kinds of candidate: those who would change the terms of building, and those who would stop approvals until new rules are written.
By our arithmetic, the roughly 30 GW in states with a pause candidate equals about half of the 62.2 GW SemiAnalysis models for 2028, against 5.7 GW if the candidates backing conditions win.
As we read SemiAnalysis’s Texas section, the timing of a restart matters: it says ERCOT is targeting 10 December for its report, which does not by itself reopen permitting.
SemiAnalysis, 1 November 2026
“The biggest risk to our buildout forecast right now is a prolonged permitting freeze at the TCEQ”
Three Implications
FOR DEVELOPERS SemiAnalysis says a pause can block any project that still needs an approval the state has stopped issuing, while conditions such as paying grid costs raise the price of building.
FOR HYPERSCALERS SemiAnalysis says several current and former hyperscaler employees it interviewed said the companies can afford higher costs from new rules, but that does not guarantee capacity arrives on time.
FOR VOTERS AND UTILITIES SemiAnalysis lists 15 candidates who want data centers to pay their own grid costs, 12 who would end tax breaks and 11 who would limit or report water use.
The 30 GW and 5.7 GW figures describe planned 2028 capacity in states where a pause would apply, not capacity that will be lost. SemiAnalysis says “How much actually slips depends on the permits each project still needs and how long the pause lasts”. It takes candidates’ positions at face value. The figures come from its own model; we read only the free part of the post. The Information’s analysis is cited from Techmeme’s summary. The 18.7 GW, the half and the eight-and-a-half comparison are our arithmetic.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
SemiAnalysis says the 2027 data center build is largely under way, and that the 2028 build of 62.2 GW is where Tuesday’s elections matter. States where pause candidates are running hold roughly 30 GW of that, against 5.7 GW if the candidates backing conditions win. Texas, with close to 30% of 2028 additions, is its biggest single risk.
A note on sourcing. The capacity figures, candidate counts and quotes come from the free part of SemiAnalysis’s 1 November post, based on its Datacenter Industry Model. The Information’s analysis is cited from Techmeme’s summary. We did not contact the candidates or SemiAnalysis. Comparisons labelled as our arithmetic are ours.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
