Big Tech claims voters’ AI backlash is “anti-Trump” not “anti-tech”

Redbud Consulting reported that 38 of 44 ballot measures across 10 states seeking to restrict data centers had passed so far, including moratoriums and bans. Analysts said the measures are unlikely to affect facilities already under construction or approved through 2027, but could constrain projects planned for 2028 and beyond. Voters also elected anti-AI candidates in key data-center states, adding to the risk of tighter restrictions on future buildout.

Analysis

The key exposure is not near-term compute demand but the marginal site: local permitting friction raises the time, legal, and power costs of projects without established approvals. That shifts bargaining power toward communities, utilities, and developers with permitted land and secured interconnections, while weakening the option value embedded in early-stage sites. Equipment suppliers may be less exposed initially if existing construction remains on schedule; the risk arrives later through cancellations or deferred orders.

The political signal is meaningful but geographically fragmented. Local measures do not establish a uniform national regime, and election promises may be diluted by implementation, litigation, or pressure to retain tax revenue and jobs. The 2028 planning horizon also makes this a weak catalyst for near-term AI revenue estimates, but a reason to discount longer-dated capacity growth and raise project-level risk premiums.

Contrarianly, a broad AI-infrastructure selloff would likely conflate unapproved projects with operating capacity and already-permitted builds. The more durable consequence may be slower, more concentrated deployment—not lower aggregate demand—benefiting incumbents able to fund compliance and secure power. No company-specific valuation or exposure data are supplied, so avoid sizing a sector short off this evidence alone.

Key Decisions for Investors

• Avoid a broad, immediate short of AI infrastructure. Treat the news as a longer-dated capacity and permitting risk, not evidence of a near-term collapse in compute demand.
• For a relative-value watchlist, favor established hyperscalers and permitted data-center operators over developers whose economics depend on unapproved sites; verify each issuer’s project pipeline, approval status, power agreements, and customer commitments before positioning.
• Over the next 1–3 months, monitor state and local implementation, permitting timelines, utility interconnection queues, and developer guidance for cancellations or delayed energization. A rise in project delays or capex deferrals would strengthen the downside thesis.
• Falsify the bearish capacity thesis if restrictions are narrowed, challenged successfully, or fail to delay projects—and if major operators maintain build schedules and capacity guidance. Reassess if utility capex plans or equipment orders start slipping.