随着人工智能的迅速发展，美国的政治格局在监管问题上变得日益分化，尽管该技术的开发者之间存在着严重的分歧，但选民们却表现出了压倒性的团结。在6月23日的国会初选中，计算机科学家亚历克斯·博雷斯以人工智能监管为竞选口号，导致反对和支持他的科技资助团体分别花费数百万美元；他最终以微弱劣势输给了米卡·拉舍。虽然政治派别担心笨拙的监管可能会让只落后几个月的中国模型赶上来，但公众的焦虑情绪仍然很高。根据民意调查，四分之三的美国人希望加强监管，近三分之二的人认为人工智能发展太快，而只有2%的人认为太慢，61%的人预计它将显著影响社会。

美国公众的担忧异常之高，46%的人预计会产生负面影响，而31%的人预计会产生正面结果。此外，45%的人对潜在的人类灭绝表示担忧，而65%的人（包括74%的民主党人和56%的共和党人）担心失去工作。超过70%的受访者担心科技公司权力集中和心理健康恶化。在进步左翼中，批评者担心工会工作被破坏并反对数据中心，同时将矛头指向科技亿万富翁。在右翼中，共和党参议员乔什·霍利警告说，人工智能可能会破坏40%至50%的入门级白领工作，这促使他支持禁止自动驾驶卡车，并警告技术精英不要进行“无限期生活在云端”的超人类主义努力。

这种广泛的社会抵制正在地方上显现，由于高耗电量和社会影响，至少有十个州引入了冻结数据中心建设的法案。联邦政策一直不一致；虽然特朗普政府此前曾试图抢先出台州级监管，但最近禁止外国人访问最强大的美国人工智能模型，导致Anthropic完全限制了访问。展望未来，提出没收主要人工智能公司50%股份给公众的左翼参议员伯尼·桑德斯和唐纳德·特朗普都同意公民应该在这些公司中拥有股份。尽管全民基本收入等细节仍存在高度争议，但政党策略师在11月选举前面临着越来越大的压力，需要制定连贯的平台来应对选民的恐惧。

As artificial intelligence advances rapidly, the political landscape in the United States is becoming increasingly divided over its regulation, even though the technology's developers are deeply split and voters show overwhelming unity. During a congressional primary on June 23rd, computer scientist Alex Bores campaigned on AI regulation, leading to opposing tech-funded groups spending millions to either defeat or support him; he ultimately lost narrowly to Micah Lasher. While political factions fear that clumsy regulation could allow Chinese models—currently only a few months behind—to catch up, public anxiety remains high. According to polling, three-quarters of Americans want more regulation, nearly two-thirds believe AI is advancing too quickly compared to just 2% who think it is too slow, and 61% expect it to affect society significantly.

American public concern is exceptionally high, with 46% expecting negative impacts compared to 31% who expect positive outcomes. Additionally, 45% express concern over potential human extinction, while 65% (including 74% of Democrats and 56% of Republicans) fear job losses. Over 70% of respondents worry about concentration of power in tech firms and worsening mental health. On the progressive left, critics worry about union job destruction and oppose data centers, while targeting tech billionaires. On the right, Republican Senator Josh Hawley warns that AI could disrupt 40% to 50% of entry-level white-collar jobs, prompting him to support banning self-driving trucks and warn against transhumanist efforts by a technological elite to "live infinitely in the cloud."

This widespread societal backlash is manifesting locally, with at least ten states introducing bills to freeze data center construction due to high electricity consumption and social impacts. Federal policy has been inconsistent; while the Trump administration previously attempted to preempt state regulations, it recently banned foreigners from accessing the most powerful American AI models, leading Anthropic to restrict access entirely. Looking ahead, both left-wing Senator Bernie Sanders—who proposed confiscating a 50% stake in major AI firms for the public—and Donald Trump agree that citizens should have a stake in these companies. While details like universal basic income remain highly contentious, party strategists are under growing pressure to develop coherent platforms addressing voters' fears before the November elections.

Source: Democrats and Republicans agree: AI is scary

Subtitle: It’s about the only thing that unites them

Dateline: 6月 25, 2026 09:14 上午 | Washington, DC