The United States is experiencing a transition from localized resistance toward AI data centers to a broader national political discourse, coinciding with the implementation of rigorous state-level AI safety legislation.

美國正經歷從針對 AI 數據中心的局部抵制，轉向更廣泛的全國性政治論述，與此同時，嚴格的州級 AI 安全立法也開始實施。

The proliferation of AI-driven data centers—exceeding 4,300 facilities nationwide—has precipitated a shift in public sentiment, with a Gallup Poll indicating a 70% objection rate among Americans. This trend is manifested in various administrative responses: New Jersey and Pennsylvania have introduced reporting requirements and benefit-demonstration standards, while municipalities in New Jersey, such as East Greenwich and Monroe, have enacted total prohibitions. In Texas, the Agriculture Commissioner has advocated for a moratorium to evaluate the impact on land and water resources. Business coalitions in New York have similarly requested a temporary cessation of permitting, citing potential electricity price inflation and environmental degradation. These frictions have resulted in the delay or cancellation of 17 and 31 projects respectively, representing a combined valuation of approximately $156 billion.

AI 驅動的數據中心大量增加——全國已超過 4,300 處設施——導致公眾情緒發生轉變，蓋洛普（Gallup）民調顯示有 70% 的美國人表示反對。這一趨勢體現在各種行政回應中：紐澤西州與賓夕法尼亞州引入了申報要求與效益證明標準，而紐澤西州的某些市鎮（如 East Greenwich 與 Monroe）則實施了全面禁令。在德州，農業專員主張暫停相關計畫，以評估對土地與水資源的影響。紐約的商業聯盟同樣要求暫停發放許可證，理由是可能導致電價上漲與環境惡化。這些摩擦導致 17 個與 31 個計畫分別被延遲或取消，總價值約 1,560 億美元。

Parallel to infrastructure concerns, a regulatory divergence has emerged between federal and state authorities. Following the federal administration's decision to rescind a vetting mechanism for frontier AI models to preserve innovation, Illinois passed SB 315. This legislation mandates that frontier AI developers submit public safety plans and undergo independent third-party audits. While firms such as OpenAI and Anthropic have expressed support—potentially to establish a uniform baseline and mitigate a fragmented regulatory landscape—trade groups like the Chamber of Progress argue that the law exposes sensitive systems to unqualified auditors. This legislative activity reflects a broader trend where states serve as laboratories for governance in the absence of comprehensive federal legislation, as noted by Illinois legislators.

在基礎設施疑慮之餘，聯邦與州政府之間也出現了監管分歧。在聯邦政府決定撤銷對前沿 AI 模型的審查機制以維護創新後，伊利諾州通過了 SB 315 法案。該立法要求前沿 AI 開發商提交公共安全計畫並接受獨立第三方審計。雖然 OpenAI 與 Anthropic 等公司表示支持——可能是為了建立統一基準並緩解碎片化的監管局面——但如進步商會（Chamber of Progress）等貿易團體則認為，該法律將敏感系統暴露給不合格的審計人員。正如伊利諾州立法者所指出的，這種立法活動反映了一個更廣泛的趨勢，即在缺乏全面聯邦立法的情況下，各州正扮演治理實驗室的角色。

The current landscape is characterized by increasing local restrictions on AI physical infrastructure and a growing trend toward state-mandated safety audits for AI developers.

目前的局面特點在於，地方對 AI 實體基礎設施的限制不斷增加，且州政府強制 AI 開發商進行安全審計的趨勢日益顯著。