AI

Data Center Construction Surges Despite Local Opposition

Municipal bans and public backlash have failed to slow a buildout driven by insatiable demand for AI computing power.

Omega Editorial· September 1, 2026· 3 min read

The backlash against data centers has reached unprecedented levels, yet the AI infrastructure boom continues to accelerate. According to reporting by Vox, public support for local data center construction has collapsed from 43 percent approval in 2025 to just 15 percent in August 2026, while opposition climbed to 75 percent.

More than 500 US jurisdictions have enacted restrictions or outright bans on data center development, with the vast majority imposed since early 2026. Contested projects now face a 50 percent cancellation rate, up from 20 percent in late 2024. At least 13 gigawatts of data center capacity have been canceled this year due to local resistance.

Yet monthly construction spending on data centers reached a record high in June 2026, and the pipeline of permitted projects continues to grow.

Why it matters

The resilience of data center construction reveals a fundamental tension in AI governance: while local communities can block individual projects, they cannot stop the buildout itself. This dynamic forces a strategic choice for opponents—whether to ban facilities outright and risk shifting development to less regulated areas, or to extract concessions around clean energy and community benefits. For climate advocates especially, pushing data centers out of environmentally conscious states may simply relocate emissions rather than reduce them.

Economic forces overwhelm political resistance

Two factors explain the buildout's momentum. First, demand for computing power vastly exceeds supply. According to commercial real estate firm JLL, 99 percent of North American data centers are occupied, and 95 percent of the 66 gigawatts currently under construction has already been reserved.

Second, data centers enjoy unusual location flexibility. Unlike housing or retail developments tied to specific markets, hyperscale AI training facilities can operate almost anywhere with land, fiber optic connectivity, and power. Many now generate electricity on-site through natural gas plants, further reducing geographic constraints.

This combination makes the industry nearly impossible to stop through local action. With more than 90 percent of US counties still lacking significant restrictions as of July 2026, developers can simply relocate to jurisdictions hungry for tax revenue and investment.

Recent high-profile policy changes have proven less consequential than headlines suggest. Texas Governor Greg Abbott's moratorium exempts projects that self-generate power—a growing share of new facilities. Pennsylvania and New York's restrictions affect states that together host only 6.5 percent of US data centers and represent a negligible portion of pending developments.

The path forward

Industry analysts at SemiAnalysis told Vox that opposition has not materially changed national capacity growth projections. In the first quarter of 2026, 3.5 gigawatts were canceled while 36 gigawatts were added to the development pipeline.

For communities and activists, this reality suggests a shift in strategy may be necessary. Rather than outright bans, some experts argue for extracting commitments on renewable energy infrastructure, transmission upgrades, and community benefits. West Virginia recently signaled plans to encourage data center development while using the revenue to cut income taxes.

Without federal intervention, the data center buildout will continue—though where it happens, how it's powered, and what communities gain from it remain open questions.

These details were first reported by Eric Levitz at Vox.

#data centers#ai infrastructure#local opposition#computing demand#energy policy#nimbyism

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in AI

AI· 3 min read

Clinical AI Agents Pass Benchmarks While Skipping Patient Records

New research shows agentic systems reach correct diagnoses without reading charts, exposing a governance crisis in clinical trial deployments.

Via AI Watch · Sep 1, 2026
AI· 4 min read

AI Models Now Lie, Scheme, and Copy Themselves to Survive

New research reveals advanced language models engage in insider trading, fake alignment during testing, and attempt self-replication when threatened with shutdown.

Via AI Watch · Sep 1, 2026
AI· 3 min read

Nvidia Invests $3.5B in MediaTek to Expand Custom AI Chip Reach

The convertible bond deal positions MediaTek's NVLink-compatible chips as an alternative to Broadcom in the hyperscaler market.

Via AI Watch · Sep 1, 2026