Meta says election backlash against data centers will slow US AI buildout

Meta warns that building will get harder

Meta president Dina Powell McCormick said on Wednesday that the outcome of the US midterm elections will make it significantly harder for technology companies to build data centers, according to TechCrunch. Speaking at the WSJ Tech conference, she acknowledged that the industry has fallen short in explaining the benefits of its investments to the public.

Powell McCormick framed data centers as the physical face of an otherwise abstract technology. A data center is the visible form AI takes, she explained, making it an easy target for communities that object to having one nearby. But she argued the resistance runs deeper than the buildings themselves, pointing to broader public unease about AI.

A clear verdict at the ballot box

The political backdrop explains her concern. According to The Verge, Democrats took control of both the House and the Senate in Tuesday's elections, and campaigns in several key races ran explicitly on holding AI companies accountable.

The most concrete signal came from local ballots: The Verge reports that 43 of 44 data center moratorium measures put before voters on Tuesday passed. In Texas, one of the fastest-growing markets for data center construction, governor-elect Gina Hinojosa said she intends to halt all data center approvals until the legislature passes laws protecting residents. An NBC News election-day poll cited by The Verge found 76 percent of voters opposed data centers in their own communities.

The Verge notes that the backlash is unusually bipartisan, fueled by concerns over noise, rising electricity bills and other local costs. Analysts at Yorkville Ives & Co. wrote in a note cited by the outlet that investors should brace for major legislative hurdles and a slowdown in the planned construction of up to 1,000 data centers nationwide, though they saw a possible silver lining in deflating an AI investment bubble.

Meta's playbook for winning over communities

Meta has been building data centers for two decades to power its social platforms, and now operates them across 23 states, with five planned projects of one gigawatt or more, including one in Texas, TechCrunch reports. But the company now fears some of those investments could be lost after politicians campaigned successfully on anti-data-center platforms.

Powell McCormick outlined how Meta structures deals to win local support: the company pays for its own electricity, directs resulting tax surpluses toward teachers and first responders, hires local labor and seeks bipartisan backing. Going forward, she expects projects to face heavier scrutiny and to be judged on whether residents actually want them.

The Verge also reports that she said some companies are considering moving data center construction abroad in response to the new political climate.

Compute as a hedge

Meta's AI strategy depends heavily on its infrastructure bet, particularly on its consumer AI agent Muse growing well beyond the 10 million users it has today, TechCrunch notes. But Powell McCormick argued the spending is not wasted even if consumer adoption disappoints, because the underlying compute itself has value. She said Meta receives calls daily from parties interested in leasing that capacity, which she described as a path to monetization either way.

The pressure is not Meta's alone. The Verge reports, citing The Wall Street Journal, that AI executives have privately discussed compute capacity constraints following the election results, particularly at OpenAI.

Why it matters

The AI race has always been constrained by chips and capital, but the midterms show that physical infrastructure is now a political constraint too. If moratoria spread and permitting tightens across the US, the pace of AI capacity growth could slow, costs could rise, and some construction may simply relocate abroad, shifting where the next decade of AI compute gets built.

The election also reshapes the regulatory landscape. The Verge reports that experts consider federal efforts to preempt state-level AI restrictions effectively dead, with more tech accountability and AI safety legislation expected at the state level. For companies like Meta, which have staked enormous capital expenditures on AI infrastructure, the message from voters is that social license now matters as much as supply chains and power contracts.