01What voters actually did
- Voters decided 41 local data-center ballot measures on election day, per Fortune citing Axios and Ballotpedia; CNBC reports they rejected development in 40 of them. Fortune adds that all six decided earlier in 2026 also went against development.
- In Texas, Governor Greg Abbott paused new approvals pending an audit, per Fortune. Texas holds about 27% of US data-center gigawattage, per Wells Fargo as cited by CNBC, and its new governor is expected to be tougher on data centers.
- In Ohio, Senator Jon Husted was defending his past support for data centers, per Fortune.
- On Thursday CNBC reported Senator Bernie Sanders called for a national data-center moratorium, and HuggingNews says some Democratic lawmakers went as far as calling for a “ban on artificial superintelligence”.
The objections in the coverage are mostly local and practical: electricity prices, water, noise, land and tax breaks. That is different from a federal ban on AI, and it is the difference that Goldman Sachs leans on.
02What can realistically change
Goldman Sachs, per Fortune, sees a national data-center moratorium as unlikely and expects the issue to fade after the election. Goldman's own figure for the exposure is different: it estimates that states likely to tighten rules hold about 45% of planned US data-center capacity, per press coverage. That does not mean 45% will be cancelled. It means that is where permits, tax breaks and electricity connections are now at risk of becoming harder to get.
UBS, per HuggingNews, estimates the outcome could put about 30% of expected 2028 US data-center capacity additions at risk, while raising its 2027 AI capital-spending forecast to $1.4 trillion. Those two facts sit together: the spending forecast is still enormous, and a large slice of its timing and location is in question. Both are analyst estimates, and we have not seen the original notes.
| Policy lever | Who holds it | Likelihood per the sources |
|---|---|---|
| Local moratoria and zoning | Counties, cities and governors | Active now; 40 of 41 measures went against development |
| State tax breaks and approvals | Governors and legislatures | Texas has paused approvals; new governors campaigned on tightening |
| Federal tax breaks for data centers | Congress | Goldman: Congress might limit them |
| National moratorium | Congress and the President | Goldman: unlikely under any outcome; Trump can veto |
| Basic federal AI oversight and testing | Congress | Goldman sees room for compromise |
03Who loses and who may gain
| Group | Reaction | Why |
|---|---|---|
| Chip makers (Nvidia, Micron, AMD) | Down roughly 8% to 15% on Thursday | Fewer or later data centers mean fewer chips ordered |
| Neoclouds (CoreWeave, Nebius, IREN) | Down 6% to 11% on Wednesday; Nebius fell double digits again Thursday | They rent AI capacity and borrow heavily to build it |
| Power suppliers (GE Vernova, Constellation) | Down about 10% and 6% on Wednesday | Data centers were the demand story |
| Hyperscalers (Meta, Alphabet, Amazon, Microsoft) | Up more than 4% on Wednesday, then sold off Thursday | Slower capex improves free cash flow, but they also run the data centers |
The weakest names are the ones that borrowed to build. Gartner predicts 80% of neoclouds could cease independent operations by the end of 2030, per CNBC, and Starwood Capital's Barry Sternlicht said some neoclouds could fail. Evercore's Julian Emanuel noted that hyperscalers, equipment providers and AI companies sought hundreds of billions of dollars of financing over the past two weeks, putting the cost of capital in focus.
04The case that this helps AI
CNBC's own headline argues the pushback “could help AI stocks”. Its reasoning, from the analysts it quotes: a slower build-out would let air out of an expensive trade gradually instead of through a crash; spenders' cash flows improve when capex pressure eases; and political compromises may still allow building in the medium term. Bank of America's Aditya Bhave called a front-loaded capex wave “symptomatic of over-investment”, and RSM's Joseph Brusuelas said “it is necessary that the investment in AI slow to a more sustainable pace”. AMD's chief executive Lisa Su, for the industry's side, said the market “can sometimes be a little bit short-sighted” and that she does not see AI chip demand slowing soon.
Both views fit the facts. Our reading is that the backlash is a constraint on location and timing rather than on demand for AI itself: models can run in data centers built elsewhere, or later, or more efficiently. What it changes is who gets paid, when, and how much debt is needed to bridge the wait. Whether the earlier spending was too much is the question in part 7.
05What it means if you build on AI
If you use AI models through an API, the practical risk is not that they stop working. It is that the price and availability of computing capacity could shift as the build-out changes pace. Avoid being locked into one provider, and keep your own records. Payaries publishes its API contract in a form AI tools can read; see Payaries for AI agents and what agentic AI is. The regulation side of the same race is in what Big Tech is asking governments for on AI.
06The full series, in reading order
This is one of eleven connected pieces on the November 2026 election and market shock. Read them in order, or jump to the one you need.
- 01The crash: what happened on 5 November
- 02How $2.7 trillion was wiped out in a day, and whether that number is real
- 03From rally to rout: why stocks rose on the Democratic win and crashed a day later
- 04The Democratic win: what the midterm result changes
- 05What a Democratic Congress means for the world economy
- 06The data-center backlash and what it means for AI (this article)
- 07Is it really an AI bubble? The evidence on both sides
- 08Gold after the midterms
- 09Bitcoin and crypto after the midterms and the chip crash
- 10What happens next: the dates, scenarios and signals to watch
- 11The discussion: the hardest questions about the crash and the AI bubble, answered
07Sources and as-of date
- Democrats may go after data centers. That's not a reason to dump all your AI stocks — CNBC
- BofA warns of AI correction risk as data-center backlash hits the 2026 midterms — Fortune
- America got the anti-data-center wave it voted for. Here come the economic consequences — CNN
- S&P 500 Wipeout Hits $2.7T as Hedge Funds Dump Hyperscalers and Chip Index Plunges 11% — HuggingNews
- Dow tumbles 1,600 points, erasing its postelection rally; Nasdaq plummets 5%: Live updates — CNBC
Questions
- What did voters decide about data centers in the 2026 midterms?
- Per CNBC, voters rejected data-center development in 40 of 41 local ballot measures. Fortune adds that all six decided earlier in 2026 also went against development.
- Will Democrats ban AI data centers?
- Senator Bernie Sanders called for a national moratorium, but Goldman Sachs sees one as unlikely under any outcome, partly because President Trump can veto it. Local and state limits on approvals and tax breaks are more likely.
- How much data-center capacity is at risk?
- UBS estimates about 30% of expected 2028 US capacity additions could be at risk; Goldman Sachs says states likely to tighten rules hold about 45% of planned capacity. Neither is a forecast of cancellations.
- Who is hurt most by data-center restrictions?
- Chip makers, debt-funded neoclouds such as CoreWeave and Nebius, and power suppliers like GE Vernova fell most. Gartner predicts 80% of neoclouds could cease independent operations by the end of 2030.
- Does this mean AI is over?
- No. The backlash concerns the buildings and the electricity, not whether AI models are useful. AMD's chief executive says she does not see AI chip demand slowing soon, though that is a company view.
Keep reading
- What a Democratic Congress and an AI sell-off mean for the world economyWhat the Democratic win and the AI-stock sell-off mean for the world economy: AI investment and growth, tariffs, oil and the Iran war, interest rates and the dollar, and how Asia and Europe reacted.
- Is it really an AI bubble? The case for, the case against, and what would settle itIs AI a bubble after the November 2026 sell-off? The evidence for (spending vs revenue, debt, circular deals, depreciation) and against (profitable spenders, real adoption), the dot-com comparison, and what would settle it.
- The US stock market crash of 5 November 2026: what happened, in numbersUS stocks had their worst day since April 2025 on 5 November 2026: the S&P 500 fell about 4%, the Nasdaq about 5% and chip stocks about 10%, erasing the post-election rally. The numbers, the names and the causes.
- What Big Tech actually wants from AI regulation: one rulebook, not noneBig Tech's 2026 AI regulation ask is one national rulebook that overrides state laws, not more rules. The White House framework, lobbying money, the EU delay.
- What is agentic AI? A plain-English guideWhat agentic AI is: AI that plans, uses tools and checks results to reach a goal. How it differs from a chatbot, how it works, the risks, and how agents pay.
- Gold after the midterms: why it climbed while stocks crashedGold rose to a two-week high near $4,130 an ounce as Democrats won Congress and stocks crashed. Why gold climbed while chip stocks fell, the role of the dollar, yields and oil, and the risks to the move.
- Bitcoin and crypto after the midterms and the chip crash: what happened and what it meansBitcoin fell about 5% from above $80,000 to roughly $76,000 on 5 November 2026 as the AI-stock sell-off hit, with hundreds of millions in liquidations. What happened, why crypto trades with tech, and what a Democratic Congress means for it.