Why Many Announced AI Data Centers May Never Get Built

A large share of the artificial intelligence data centers now being announced by developers and utilities may never reach construction, according to Brent Nelson, senior managing director of markets and strategy at Ascend Analytics, who discussed the gap between announcement and completion on TechCrunch's Equity podcast on Nov. 11, 2026. The episode, hosted by senior reporters Rebecca Bellan and Tim De Chant, set out to separate data center projects "that are likely to get built from the ones that may never get off the ground," according to TechCrunch. (TechCrunch)

The distinction matters because, as TechCrunch framed it, billions of dollars in AI infrastructure investment are tied to projects whose completion is not guaranteed. Before a facility can be judged against permitting timelines, grid capacity or community resistance, the more basic question is whether it is a committed build or a speculative announcement, the outlet reported.

What Makes a Data Center Project Speculative Rather Than Real?

According to TechCrunch's reporting, the data center boom is colliding with three recurring obstacles before ground is ever broken: local opposition, permitting delays and a scramble for electricity supply. Those three factors function less as abstract risks and more as filters — mechanisms that determine which announced megaprojects convert into physical buildings and which remain press releases. Nelson's role at Ascend Analytics, a markets-and-strategy advisory firm, is to help clients and the podcast's audience sort projects along exactly that line, the episode's description states.

TechCrunch did not publish specific figures on how many announced projects fall into each category, and this report includes only the obstacles the outlet named rather than estimates not present in the source material.

Why Does Local Opposition Slow or Stop Projects?

TechCrunch describes "widespread local opposition" as one of the real-world obstacles now confronting the data center build-out. In practice, opposition at the community or municipal level can delay or derail a project long before a shovel goes in the ground, since local governments typically control zoning approvals and land-use permits that a developer needs before construction financing closes. The podcast did not name specific towns, companies or projects facing opposition, and this article does not attribute opposition to any named community beyond what TechCrunch reported in general terms.

How Do Permitting Delays Filter Out Weaker Proposals?

Permitting delays are listed alongside local opposition as a core obstacle in TechCrunch's framing of the sector. Permitting functions as a gating mechanism: a project that cannot clear environmental review, utility interconnection studies or local building approvals within a developer's financing window risks losing the capital commitments that made the announcement possible in the first place. That dynamic, as described in the podcast, is part of why the number of data centers announced in press releases and the number that ultimately get built can diverge.

What Role Does the Electricity Supply Play?

The third obstacle TechCrunch cites is a "scramble for electricity" — a reference to the competition among data center developers, utilities and other large power users for available grid capacity. Because AI-focused data centers require substantial, reliable electricity supply, a project without a secured power contract or grid interconnection is, by definition, incomplete, regardless of how far along its site plan or financing otherwise is. TechCrunch's episode positions this power question as one of the clearest dividing lines between projects likely to be built and those that may stall.

Who Is Assessing Which Projects Will Survive?

The analysis comes from Brent Nelson of Ascend Analytics, speaking on Equity, TechCrunch's flagship podcast, in an episode hosted by Rebecca Bellan, who covers the business and policy of artificial intelligence, and Tim De Chant, the outlet's senior climate reporter. The episode runs roughly 31 minutes and is distributed on YouTube, Apple Podcasts, Overcast and Spotify, according to TechCrunch's posting. Audio producer Theresa Loconsolo is credited with the episode's production. No additional named sources, officials or company representatives are quoted in the source material reviewed for this report.

What Is at Stake for AI Infrastructure Spending?

TechCrunch frames the stakes in financial terms: with billions of dollars in AI infrastructure investment riding on these announcements, the outlet reported that "separating realistic projects from speculative ones is becoming increasingly important" for investors, utilities and policymakers tracking the sector. The podcast did not provide a specific investment total, project count or completion-rate figure, and none is included here because it was not present in the source material. For readers following the broader AI infrastructure build-out, the episode's core point is procedural rather than numerical: a data center announcement is not evidence of a data center, and the obstacles standing between the two — opposition, permits and power — are the variables analysts like Nelson say determine which projects ultimately get poured in concrete.

mindfulAI Keyboard — An iPhone keyboard that rewrites a draft in a tone you pick. Four free AI actions a day.