The American technology boom is acquiring a more visible political risk. Following Democratic victories in the midterms, investors reassessed AI exposure as concerns about regulation and opposition to data centres contributed to pressure on the dollar. That reaction reflects expectations about future policy. It does not establish that Congress has enacted new restrictions or that the investment cycle has already reversed.
The deeper issue is that AI expansion depends on infrastructure with consequences far beyond software companies. The International Energy Agency projects electricity demand from data centres worldwide to reach about 945 terawatt hours in 2030 in its base case, more than twice the 2024 level. That estimate covers data centres generally, not AI alone. AI is an important driver, but conflating all digital electricity consumption with AI would exaggerate its individual footprint.
Those requirements create legitimate distributional questions. A project can benefit investors and users while leaving nearby communities to debate electricity supply, infrastructure expansion and the allocation of costs. The relevant policy test is whether developers demonstrate credible arrangements for meeting demand and paying for associated obligations. Public scrutiny need not amount to opposition to technological progress; it can determine whether expansion remains politically and economically durable.
Competition policy adds another source of uncertainty. Brookings anticipates greater scrutiny of technology platforms and the possibility of concentrated power in AI markets. Investigations could ask whether established firms use their advantages to restrict meaningful competition. That inquiry should remain distinct from assuming that every large investment or commercial partnership is anticompetitive.
Federal legislation still faces institutional constraints. The president can veto legislation, with a two thirds vote in each chamber required for an override. An incoming congressional majority therefore changes negotiating conditions without guaranteeing a particular regulatory outcome. Markets may react before policymakers agree on what they intend to do.
For the AI industry, the practical implication is to make expansion easier to defend publicly. Transparent infrastructure costs, credible competition and evidence of useful applications provide stronger foundations than promises of inevitable economic transformation. For policymakers, the challenge is to distinguish productive oversight from unpredictable intervention. The next phase of AI investment will test whether commercial ambition can coexist with institutions capable of examining its wider costs.
Original analysis inspired by Li Gu and Niket Nishant from Reuters. Additional research and verification conducted through multiple sources.
