Story
Nov 9, 2026

Data-center backlash turns AI’s trillion-dollar buildout into a political risk

Voters and newly empowered Democrats have signaled that AI infrastructure cannot expand without local consent and tougher rules, while industry leaders and investors warn that those constraints could slow a vast, economically important buildout rather than stop it outright.

Tuesday’s midterm elections turned a long-running local fight over power-hungry data centers into an immediate Wall Street concern. Voters backed restrictions in several states, including 21 anti-data-center measures in Ohio; Piqua’s vote could jeopardize Meta’s planned $1 billion facility. Texas Gov.-elect Gina Hinojosa and Georgia’s incoming governor, Keisha Lance Bottoms, also supported moratoriums while regulation catches up.

The message was not simply anti-technology. AMD chief executive Lisa Su argued Wednesday that the sector has to make its case locally: “We need to make sure that the value proposition is clear to every local community that is going to host a data center.” Meta President Dina Powell McCormick, however, acknowledged the commercial consequence: building U.S. data centers is “going to be harder,” and hyperscalers may hedge by considering projects abroad.

By Thursday, that political uncertainty collided with an AI trade already strained by high rates, heavy capital spending and doubts over returns. The Nasdaq fell about 5.5% and the S&P 500 roughly 4%, erasing much of the prior day’s post-election gains. Nvidia dropped 9.3%, AMD 17.3%, Micron 17.5% and Broadcom 10.4%, while debt-heavy neocloud operators also sold off sharply.

The anxiety spilled into Silicon Valley’s own political conversation. David Sacks amplified a podcast framing the rout as a “$4T Wipeout” driven by “Data Center Fears” and the prospect of an AI ban.

Analysts at Yorkville Ives now expect Big Tech capital-spending growth to peak in 2027, earlier than previously forecast, and estimate at least 10% of the 800 to 1,000 planned U.S. AI-focused data centers could be removed from roadmaps. Yet Barclays strategist Dominique Toublan offered a less apocalyptic reading: local pushback may make construction costlier, not impossible—turning the AI race into a “war of attrition” in which scarce power rewards the deepest balance sheets.

By Friday morning, tech shares had begun to recover. But the market had absorbed a new reality: the decisive bottleneck for AI may no longer be chips alone, but public permission to build the places that run them.

Story coverage