Tom's Hardware reports that US midterm elections have produced significant local opposition to AI infrastructure, with voters approving numerous measures to limit data center construction.reported
Evidence: reported
AI data centers comprehensively defeated at US midterm electionsaifleet-tomshardware
The claim
According to Tom's Hardware, early election results show Americans are overwhelmingly voting against data center developments.reported The outlet cites Data Centers on the Ballot, which tracks 33 measures restricting development that have already won, alongside six more questions that are leading.reported
Evidence: reported
AI data centers comprehensively defeated at US midterm electionsaifleet-tomshardware
Evidence: reported
AI data centers comprehensively defeated at US midterm electionsaifleet-tomshardware
The response
The source notes that only one measure in Harwood, North Dakota, which adds an extra review process for industrial projects, has failed. A second tracker of 46 measures across ten states indicates that 11 of 13 positive measures to permit data centers look set to be defeated.reported
Evidence: reported
AI data centers comprehensively defeated at US midterm electionsaifleet-tomshardware
The practical upshot
Despite the voting trends, the material states it is unclear if these results will affect projects already underway. Out of more than 300 local bans, only three developments totaling 2.3 gigawatts have been delayed, while 38 gigawatts are expected to come online by 2027.
Claims versus evidence
The headline assertion of a comprehensive defeat contrasts with the evidence that existing infrastructure plans remain largely intact. Tom's Hardware reports that investment is still on track to hit $32 trillion by 2050.reported The source acknowledges the limit of these votes: they may affect future developments, but current builds are proceeding. OpenAI and Anthropic are scrambling for small data center projects ranging from 20 to 30 megawatts, especially as their hyperscale gigawatt sites face delays, according to the report.reported This suggests a strategic shift toward smaller footprints rather than a total halt. The U.S. Senate accuses Big Tech of misleading the public on the advantages of AI data centers, adding political pressure to the operational challenges.
Evidence: reported
AI data centers comprehensively defeated at US midterm electionsaifleet-tomshardware
Why it matters
- Product teams must account for potential delays in hyperscale infrastructure deployment, as noted by Tom's Hardware.
- The pivot to smaller 20-to-30 megawatt projects by major AI firms indicates a changing supply dynamic for compute resources.
- Local opposition trends may influence long-term capacity planning and regional deployment strategies for AI workloads.
- Political scrutiny, including Senate accusations, adds regulatory risk to future infrastructure announcements.
Topics: AI infrastructure, data centers, OpenAI, Anthropic, US elections
Updated
Then / now / next
AI firms pursued large-scale gigawatt data center projects with varying levels of local support.
Local voters have passed 33 restrictive measures, prompting a shift toward smaller infrastructure projects.
Watch for delays in hyperscale sites and the impact of Senate scrutiny on future approvals.
Story timeline
- Then AI firms pursued large-scale gigawatt data center projects with varying levels of local support.
- 04 Nov 2026 AI data centers comprehensively defeated at US midterm elections
- Now Local voters have passed 33 restrictive measures, prompting a shift toward smaller infrastructure projects.
- Next Watch for delays in hyperscale sites and the impact of Senate scrutiny on future approvals.
The strongest doubt
Counterpoint from the sources
The source notes that 38 gigawatts are expected to come online by 2027 and investment remains on track, suggesting the election results may not significantly alter near-term infrastructure availability.
For readers
Developers
- Expect potential delays in large-scale compute availability as firms pivot to smaller sites.
Business
- Monitor the shift toward 20-to-30 megawatt projects by major AI vendors like OpenAI and Anthropic.
Policy
- Local opposition remains strong, with 33 restrictive measures passing in recent elections.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Questions readers ask
How many data center restrictions passed in the midterms?
Tom's Hardware reports that 33 measures restricting data center development have won, with six more leading.
Are large AI data center projects being cancelled?
No, the source states it is unclear if votes affect underway projects, with 38 gigawatts expected by 2027.
Why are OpenAI and Anthropic looking at smaller sites?
They are scrambling for smaller projects as their hyperscale gigawatt sites face delays, according to the published report
Connected stories
Reading receipt
- Tom's Hardware reports 33 ballot measures restricting data centers have won, with six more leading.
- OpenAI and Anthropic are pursuing smaller projects of 20 to 30 megawatts amid delays at larger sites.
- Investment in AI infrastructure is projected to reach $32 trillion by 2050 despite the election results.
Still open: Will the Senate accusations lead to federal regulatory changes affecting data center approvals?
Still open: How will the delay of hyperscale sites impact the timeline for the 38 gigawatts expected by 2027?