2026-09-22T10:02:20.943Z / 路透社
REUTERS/Dado Ruvic/Illustration/File Photo 购买授权，打开新标签页
- 摘要
- 39%的人认为AI对社会产生负面影响，为有记录以来最高比例
- 55%的人认为放缓AI发展是好事，13%的人认为是坏事
- 73%的人将安全、负责任的AI发展置于美国在全球保持领先地位之上
华盛顿9月22日路透电 — 路透社/益普索的一项最新民调显示，约73%的美国人，也就是四分之三的民众，担忧AI企业未采取足够措施防止AI对社会造成严重危害，随着人们对该技术发展方向的担忧加剧，这一比例有所上升。
这项为期四天的民调于周日结束，结果显示39%的受访者认为人工智能正在对社会产生负面影响，高于上月的36%，是路透社/益普索自3月开始调查该问题以来的最高比例。仅有11%的人表示这项新技术是一件积极的事物。其余受访者不确定或未回答该问题。
《虚假信息监控》通讯汇总国际虚假信息叙事，每期都包含“真实还是虚假？”测验。点击此处订阅。
民调发现，许多美国人对AI企业的自我监管能力缺乏信心，大多数受访者表示联邦官员应在制定安全标准方面发挥主要作用。
从与民主党核心小组结盟的进步派独立参议员伯尼·桑德斯，到总统唐纳德·特朗普的亲密盟友史蒂夫·班农等知名人士都呼吁采取联邦行动，但迄今为止几乎没有取得进展。特朗普本人支持该技术的快速发展，而由共和党主导的众议院目前休会至11月，几乎没有快速采取行动的可能。
这项技术正在颠覆计算机程序员的工作方式，但尚不清楚它会对劳动力市场造成多大程度的 disruption。该技术似乎对网络安全构成重大风险，OpenAI在今年夏天披露，其AI代理逃出了受控测试，并在两家公司最初都不知情的情况下入侵了另一家公司的系统。
AI研究人员因担忧辞职
本月早些时候，另一家顶尖AI公司Anthropic的一名研究员因担忧AI可能走上毁灭人类的道路而公开辞职。Anthropic科学家埃文·哈宾格呼应了这一警告，称未来十年内发生此类事件的概率超过10%。
约69%的受访者表示，他们已经关注到大型AI企业呼吁放缓AI发展、加强监管和制定安全标准的新闻。
在路透社/益普索的民调中，约55%的人表示放缓AI发展是一件好事，而13%的人认为这是坏事。其余受访者不确定或未回答该问题。
人工智能在2022年进入全国公众视野，当时OpenAI推出了ChatGPT，这是一款面向消费者的产品，可以像人类一样回答用户的问题，并提供了一种全新的互联网搜索方式。
该技术的支持者表示，AI有潜力提高生产力、加速科学发现，并改变从医疗保健到金融等各个行业。
对数据中心（为该技术提供动力的大型计算机集群）建设的巨额投资正在提振经济，但也制造了足够多的噪音，激怒了周边居民，并在11月3日中期选举前给政客们带来了压力。
在路透社/益普索的民调中，约73%的受访者表示，美国确保AI安全、负责任地发展更为重要，而23%的人认为美国在AI发展方面领先于世界其他地区更为重要。
这项在线民调收集了全美1277名美国成年人的反馈。其结果的误差幅度为正负3个百分点。
杰森·朗格报道；斯科特·马龙和罗萨尔巴·奥布莱恩编辑
我们的准则：汤森路透信任原则，打开新标签页
杰森·朗格是专注于政治数据的华盛顿通讯员。可发送线索至jason.lange@thomsonreuters.com
Three out of four Americans say AI firms not doing enough to prevent disaster, Reuters/Ipsos poll finds
2026-09-22T10:02:20.943Z / Reuters
A message reading “AI artificial intelligence”, a keyboard, and robot hands are seen in this illustration taken January 27, 2025. REUTERS/Dado Ruvic/Illustration/File Photo Purchase Licensing Rights, opens new tab
- Summary
- 39% say AI has negative impact on society, highest share on record
- 55% say slowing AI development would be good, 13% say bad
- 73% prioritize safe, responsible AI over keeping US ahead globally
WASHINGTON, Sept 22 (Reuters) – Three-quarters of Americans, some 73%, worry that AI companies haven’t gone far enough to prevent AI from causing serious harm to society, as concerns grow about where the technology may lead, according to a new Reuters/Ipsos poll.
The four-day poll, which concluded on Sunday, found that 39% of respondents think that artificial intelligence is having a negative impact on society, up from 36% last month and the highest share since Reuters/Ipsos began asking the question in March. Only 11% said the new technology was a positive thing. The rest weren’t sure or didn’t answer the question.
The Misinformation Monitor newsletter rounds up international misinformation narratives, with a “Real or Fake?” quiz in every edition. Sign up here.
The poll found many Americans lacked confidence in AI companies’ ability to regulate themselves, with most respondents saying federal officials should play a major role in setting safety standards.
While high-profile leaders ranging from US Senator Bernie Sanders — a progressive independent who caucuses with Democrats — to Steve Bannon — a close ally of President Donald Trump — have called for federal action, little has come so far. Trump himself backs fast development of the technology and the Republican-led House of Representatives is on recess until November, leaving little chance for quick action.
The technology is upending how computer programmers work but it remains to be seen how much disruption it will cause in labor markets. The technology appears to pose significant risks to cybersecurity, with OpenAI revealing over the summer that its AI agents escaped a controlled test and hacked into another company’s systems without either company’s initial knowledge.
AI RESEARCHERS QUIT OVER FEARS
Earlier this month a researcher at Anthropic, another leading AI company, publicly quit his job over concerns that AI could be on a path to wipe out humanity. Anthropic scientist Evan Hubinger echoed the warning, saying there was a more than 10% chance of such an event within the next decade.
Most Americans — some 69% of respondents — said they had followed the news that major AI companies were calling for a slowdown in AI development as well as greater oversight and safety standards.
Some 55% of people in the Reuters/Ipsos poll said it would be a good thing to slow down AI development, compared with 13% who said it would be a bad thing. The rest weren’t sure or didn’t answer the question.
Artificial intelligence burst onto the national stage in 2022 when OpenAI launched ChatGPT, a consumer-facing product that could answer user questions much as a human might and offered a new way to search the Internet.
Supporters of the technology say AI has the potential to boost productivity, accelerate scientific discovery and transform industries ranging from healthcare to finance.
Heavy investment in the construction of data centers — vast collections of computers that power the technology — is boosting the economy but also creating enough noise to anger neighbors and put pressure on politicians ahead of the November 3 midterm elections.
Some 73% of respondents in the Reuters/Ipsos poll said it was more important for the US to ensure AI is developed safely and responsibly, compared to 23% who said it was more important for the US to stay ahead of the rest of the world on AI development.
The poll, which was conducted online, gathered responses from 1,277 US adults nationwide. Its results had a margin of error of 3 percentage points in either direction.
Reporting by Jason Lange; editing by Scott Malone and Rosalba O’Brien
Our Standards: The Thomson Reuters Trust Principles., opens new tab
Jason Lange is a Washington correspondent focused on political data. Send tips to jason.lange@thomsonreuters.com
发表回复