Thursday, September 17, 2026
AI 인프라 · 뉴스 & 분석
정책리포트
정책 · 리포트

A proposed data center bill aims to shield residential consumers from rising electricity costs driven by AI infrastructure expansion.

Legislative intervention could impose rate structures or caps that affect hyperscaler utility contracts, forcing developers to negotiate more favorable tariff classifications or invest in dedicated generation.
업계 전문지Slicast · 2026년 9월 16일 15:16 UTC · 미국 · 출처: The Independent
중요도 65

The U.S. House of Representatives is set to vote on the bipartisan Ratepayer Protection Act, legislation aimed at shielding American households from soaring electricity costs driven by the rapid expansion of artificial intelligence data centers. Introduced by Democratic Representative Kathy Castor and Republican Representative Gabe Evans, the bill requires state utility regulators to evaluate whether high-capacity energy users, including data facilities, should directly fund the electrical grid infrastructure built to serve them.

House Democratic Leader Hakeem Jeffries endorsed the proposal as an appropriate step. Former FERC member Allison Clements highlighted that the bill’s bipartisan support sends a clear signal that Congress prioritizes ratepayer protections amidst rising energy demands.

Environmental organization Food & Water Watch criticized the legislation, stating it focuses strictly on power expenses while overlooking broader ecological and community impacts, such as massive water consumption and pollution.

Public sentiment strongly aligns with the bill’s intent. A new AP-NORC/EPIC poll shows that about seven in 10 Americans support forcing data center companies to pay for necessary grid upgrades, with a majority also favoring clean energy requirements and construction limits. The scale of these developments underscores the urgency of the policy debate; for instance, the Stargate Oracle AI data center campus in Abilene, Texas, is projected to span roughly 4 million square feet.

원문 보기
A proposed data center bill aims to shield… · Slicast