Voters reject data center proposals, threatening AI infrastructure expansion
Forty-one of forty-two anti-data center measures passed in U.S. ballots this week, marking the first time regular Americans have wielded significant leverage over Silicon Valley infrastructure plans. The rejection threatens the business models of levered AI compute companies that rely on incremental expansion.
Over 40 data center proposals appeared on U.S. ballots this week, and 41 of 42 anti-data center measures passed. This outcome represents the first time regular Americans have wielded significant leverage over Silicon Valley infrastructure plans. The rejection is attributed to the manner in which data centers were rolled out, often involving non-disclosure agreements, a lack of community engagement, and accusations of corporate overreach by figures such as Elon Musk and Kevin O'Leary.
The backlash threatens the business models of levered AI compute companies like Anthropic, OpenAI, and CoreWeave, which rely on incremental compute expansion. Scott Galloway argued that while regulation feels good, it often acts as a moat for incumbents like Amazon and Microsoft who already possess compute, while hurting new entrants. He suggested that proper data center deployment should include closed-loop water systems, no non-disclosure agreements, guaranteed property taxes, and community benefits, rather than being decided by a small group of AI executives. Galloway characterized the voter rejection as a pivotal shift in power dynamics, forcing Silicon Valley to seek permission from the public and potentially slowing the AI infrastructure buildout.