返回新聞
政策AI Understanding 簡報

《連線》報道候選人簽署人工智慧協議,將資料中心限制與模型安全連結起來

据《连线》报道,超过 15 名候选人签署了一项五点人工智能协议,涵盖数据中心监管、强制性模型安全审查、工人福利和法律责任。該報告的說法尚未在這裡得到獨立證實。

6 min readRead the original reporting
Source-provided image accompanying WIRED reports candidates sign AI pact linking data-center limits to model safety
歸因報告來源記錄
出版商
wired.com
來源連結
wired.comhttps://www.wired.com/story/candidates-are-signing-a-pact-promising-action-on-data-centers-and-ai-safety/
來源類型
新聞媒體的報道-不是第一方文件。
還引用了

我們無法獨立確認的內容: 此聲明歸因於指定的商店。我們沒有根據第一方文件對其進行驗證。 (wired.com)

故事最後修訂

背景60 秒內了解這一點

從這裡開始

關鍵術語

人工智慧(AI)
建構執行需要模式識別、推理、語言或決策的任務的系統的廣泛領域。
人工智慧安全
該領域專注於減少人工智慧系統中的有害行為、故障和誤用風險。
測試一下自己什麼是人工智慧?測驗

自發布以來發生了什麼變化

  1. 首次發表
  2. WIRED reports additional detail on the same AI Pact event: more than 15 candidates have signed five commitments covering mandatory AI-model safety reviews, an AI dividend for workers, legal recourse over model harms, and limits on data-center subsidies and energy use. The report also adds candidate and organizer comments about the pact’s political strategy and its relationship to labor and state data-center debates.

發生了什麼事

WIRED reports that more than 15 politicians running for office have signed the AI Pact, a pledge combining data-center restrictions with policies addressing and economic disruption. The pact calls for mandatory safety reviews for AI models, an AI dividend for workers, and legal rights for consumers and companies harmed by models.

WIRED reports that more than 15 candidates across the United States have signed the AI Pact, a five-promise platform focused on both data centers and artificial intelligence. Nebraska independent Senate candidate Dan Osborn is identified as one of the signers. According to WIRED, all current signers except Osborn are Democrats, while the organizers say Republican candidates are also welcome to join. The report does not provide a complete list of signers in the supplied text, so the exact number and composition of the group remain unconfirmed here.

The pact was developed by the Political Integrity Project, an anti-money-in-politics group. WIRED reports that cofounder Daniel Lobo-Lewis said the organizers worked with groups and political organizations involved in AI legislation when creating the five promises. The commitments include support for mandatory safety reviews of AI models, policies that would establish an AI dividend for workers, and protections allowing consumers and companies to sue over harms caused by models.

WIRED reports that the pact does not explicitly demand a moratorium on data-center construction. Instead, it calls for policies that would ban tax breaks, backroom deals, and excessive energy use associated with data centers. Lobo-Lewis described that language as a baseline or floor intended to be acceptable to candidates with ties to labor. The report says several labor unions have recently reaffirmed support for data-center construction, creating a political tension between demands for infrastructure and concerns about its costs.

The report places the pact within a broader political response to public concerns about data centers. WIRED says candidates are responding to worries about environmental impacts, construction secrecy, corporate influence, land loss, and electricity use. It cites recent actions by Pennsylvania Governor Josh Shapiro and Texas Governor Greg Abbott, both of whom have taken steps to reassure voters that data-center development is being regulated. Abbott, according to WIRED, initiated a temporary construction moratorium in Texas this month. These surrounding developments are reported context; the supplied source does not independently establish the effectiveness or legal status of any of the measures.

來源詳情: wired.com ↗

為什麼這很重要

The pact reflects a growing effort to connect the local effects of data-center construction with national debates over AI accountability. It also shows candidates attempting to define a minimum policy position that can attract support across districts where voters are weighing jobs, energy use, land development, privacy, and the potential effects of automation.

The pact matters because it treats AI policy and physical infrastructure as connected political questions. Data centers are where computational demand becomes visible through land use, electricity consumption, construction, and local economic effects. WIRED reports that candidates are using those tangible concerns to discuss less visible consequences of AI, including potential job disruption, privacy risks, and the accountability of companies deploying models.

The proposed safety reviews would represent a significant change from relying primarily on voluntary commitments by AI companies. WIRED reports that the White House is developing a voluntary safety review for AI models, while the AI Pact advocates mandatory reviews. The difference is consequential: a voluntary process may depend on company participation and executive policy, whereas a mandatory process would require a legal authority, defined standards, enforcement, and resources. The source does not specify how the pact’s proposed reviews would work or which agency would administer them.

The worker-dividend proposal also links AI deployment to distributional questions. WIRED reports that former President Donald Trump has discussed the idea of a worker dividend for AI profits after it was reportedly pitched by OpenAI CEO Sam Altman. The pact presents the concept as a policy commitment, but the source gives no details about eligibility, funding, payment levels, or whether the proposal would be financed by taxes, company contributions, or another mechanism. Its practical impact therefore cannot be assessed from this report alone.

Legal accountability is another central element. WIRED reports that the pact would preserve the ability of consumers and companies to sue over harms caused by models, and that Texas candidate Gina Hinojosa has argued that data centers supporting an AI industry without rules threaten jobs, health, safety, and well-being. Such claims identify potential public stakes, but the source does not document a specific model-related injury or court ruling arising from the pact. It also does not establish whether existing legal protections are being removed or what new liability rules signers would support.

Interactive Mechanism

互動機制:它實際上是如何運作的

以互動方式探索這項發展背後的基礎技術。

System Requirements:
Best ArchitecturePure RAGRecommended pattern
Hallucination RiskVery LowGrounding efficacy
Update Cost$0 (Vector sync)Ongoing maintenance
Core takeaway: Fine-tuning teaches models how to speak (form, style, syntax); RAG teaches models what to say (verifiable facts). Never use fine-tuning alone for factual memory.
互動式概念檢查+10 Points
What is AI? Quiz

A route planner searches possible journeys using explicit rules. What does this illustrate about AI?

接下來看什麼

The key test is whether signers turn the pledge into legislation, campaign commitments, or oversight proposals. Important unknowns include the full list of signers, the pact’s precise enforcement mechanisms, whether any federal or state safety-review system will be created, and whether data-center limits will gain support from labor groups and voters.

The immediate question is whether the AI Pact grows beyond its current group of candidates. WIRED reports that the organizers welcome Republican signers, but the supplied report says the current signers are overwhelmingly Democratic. A broader coalition could make the pledge more politically durable, while a narrow coalition could limit its influence in districts where data-center construction is promoted as an economic-development strategy.

Observers should also distinguish campaign language from enacted policy. The pact’s promises cover mandatory model reviews, worker payments, litigation rights, and restrictions on subsidies and energy use, but WIRED does not report that any of these commitments has become law. The source also does not identify legislative text, proposed enforcement rules, deadlines, or official estimates of costs. Whether candidates support the same definitions of excessive energy use, model harm, or an AI dividend remains unknown.

Data-center policy is likely to remain a contested issue within labor and economic debates. WIRED reports that candidates see possible construction jobs and infrastructure benefits, while some voters and campaigners fear land loss, higher costs, pollution, and longer-term job displacement. Osborn’s comments, as reported by WIRED, illustrate that tension: he said data centers can provide infrastructure and jobs but also said a pause in construction would not be unreasonable. The source provides no independent polling detailed enough to determine how voters weigh those tradeoffs.

The relationship between public concern about data centers and concern about AI itself is another unresolved issue. WIRED reports that some polling suggests environmental concerns about data centers outrank AI concerns, and that commentators question how closely the two issues are connected. The report also mentions privacy concerns and the spread of Flock cameras in Osborn’s remarks, but it does not establish a causal link between those concerns and the AI Pact. Further reporting would be needed to verify the pact’s full membership, track concrete legislative actions, and assess whether its proposed safeguards produce measurable protections.

相關指引和測驗

什麼是人工智慧?人工智慧模型解釋AI 倫理AI 的未來測試你所知道的—嘗試免費的人工智慧測驗在我們的詞彙表中尋找人工智慧術語關注AI監管追蹤器

更新和更正

當正在發生的事件發生重大變化時,這個典型的故事就會被更新。它的 URL 和原始發布日期永遠不會改變。

  • WIRED reports additional detail on the same AI Pact event: more than 15 candidates have signed five commitments covering mandatory AI-model safety reviews, an AI dividend for workers, legal recourse over model harms, and limits on data-center subsidies and energy use. The report also adds candidate and organizer comments about the pact’s political strategy and its relationship to labor and state data-center debates.
查看公開更正日誌
覺得有用嗎?