字号·· | 护眼
半岛电视台

美国非营利组织起诉科技巨头,要求叫停AI开发US nonprofits sue tech giants to halt AI development

点「原文对照」整页切到原文,或双击某段只看那段的原文。

人工智能保障措施,包括人工控制,是一起针对全球最大科技公司的诉讼的核心[资料图:帕特里克·西森/美联社照片]Artificial intelligence safeguards, including human controls, are at the centre of a lawsuit against the world's largest tech companies [File: Patrick Sison/AP Photo]

一个由数字安全和消费者监督机构组成的联盟对全球顶尖人工智能(AI)开发商提起诉讼,请求法院下令冻结AI模型的训练和部署。

A coalition of digital safety and consumer watchdogs has filed a lawsuit against the world’s top artificial intelligence (AI) developers, requesting a court-ordered freeze on AI model training and deployments.

这起民事诉讼由不可替代教育基金、公共公民组织、热度倡议和安全AI联盟于周四向旧金山高等法院提起。此前,民主党在中期选举中取得压倒性胜利。许多分析人士指出,这一结果可能重塑并最终收紧对硅谷的监管框架。

The civil complaint, filed Thursday in San Francisco Superior Court by the Irreplaceable Education Fund, Public Citizen, Heat Initiative and the Alliance for Secure AI, comes on the heels of a sweeping Democratic midterm victory that many analysts note could reshape, and ultimately tighten the regulatory framework for Silicon Valley.

诉讼针对五家科技巨头——Anthropic、谷歌、Meta、OpenAI和SpaceX——指控这些行业领军企业无法阻止其AI产品造成“灾难性伤害”。原告要求这些公司在落实安全保障措施之前暂停相关技术。

The suit targets five tech giants – Anthropic, Google, Meta, OpenAI and SpaceX – alleging the industry leaders cannot prevent their AI creations from causing “catastrophic harm”. They are demanding the companies pause their technologies until safeguards are put in place.

诉讼援引了有记录在案的AI驱动网络攻击及其他有害行为案例,并指出这些公司自己也曾郑重承认,其未受监管的技术可能失控,对人类构成生存威胁。

The lawsuit cites documented instances of AI-driven cyberattacks and other harmful behaviours, alongside the companies’ own grave admissions that their unregulated technologies could go rogue and pose an existential threat to humanity.

这些非营利组织寻求对大型科技公司的下一代模型发布紧急禁令。他们要求,这些公司只有在能够证明技术处于人类控制之下时,才被允许恢复运营。

The nonprofits are seeking an immediate emergency injunction on next-generation models from major tech firms. They are demanding the companies only be allowed to resume operations once they can prove the technology is under human control.

“没有任何其他行业被允许将其自身开发不安全产品——更不用说可能杀死我们所有人的产品——视为不可避免,”起诉书写道。“原告提起本诉讼,是为了阻止被告继续沿着违法道路走向灾难。”原告声称,这些公司危及公共健康并干扰计算机系统等财产的使用,违反了加州的不正当竞争法。该法律禁止欺骗性广告、反竞争市场行为以及诱饵调包手法。

“No other industry is allowed to treat its own development of unsafe products — much less ones that could kill us all — as inevitable,” reads the complaint. “Plaintiffs bring this suit to stop Defendants from continuing their unlawful path to disaster.”The plaintiffs claim that the companies have violated a California unfair competition law by endangering public health and interfering with the use of property, such as computer systems. The law prohibits deceptive advertising, anti-competitive market behaviours and bait-and-switch tactics.

这些倡导组织还表示,不得不挪用资源来应对这些公司快速推进的AI模型所带来的“近乎每日发生的危机”。

The advocacy groups also report having to divert resources to counter “near-daily crises” from the companies’ rapidly advancing AI models.

将法院下令的临时暂停变为永久禁令,会带来一些棘手问题:如何执行禁令、如何核实企业是否遵守,以及最终在解除冻结前需要达到哪些法律门槛。

Turning a temporary court-ordered pause into a permanent ban raises tough questions about how to enforce the ban, how to verify that companies are following it, and ultimately, what legal thresholds are required before the freeze could be lifted.

四名原告之一“不可替代教育基金”告诉半岛电视台,该组织要求政府对人工智能开发进行监管。这家总部位于纽约的非营利倡导组织专注于为公众提供技术保障。

One of the four plaintiffs – the Irreplaceable Education Fund, a New York-based nonprofit advocacy organisation focused on technology safeguarding for the public – told Al Jazeera it is demanding government regulation of AI development.

“真正的冻结必须明确哪些工作必须停止、如何核实合规,以及违反规则的后果。”“不可替代”创始人兼执行主任菲尔·阿罗内亚努说。

“A real freeze must specify what work has to halt, how compliance is verified, and the consequences of breaking the rules,” said Phil Aroneanu, founder and executive director of the Irreplaceable.

“这需要政府检查和叫停被禁止工作的权力,而不仅仅是对公开发布模型实施禁令。为帮助执行这一冻结,可以对大型计算设施进行监控。”在被追问希望建立哪些具体法定保障措施时,该组织概述了一套多层级监督体系。

“This requires government inspections and the authority to halt prohibited work, not just a ban on publicly releasing the models. To help enforce this freeze, major computing facilities can be monitored.”Pressed on what specific statutory guardrails they hope to establish, the group outlined a multi-tiered system of oversight.

这家非营利组织表示,人工智能开发应以“经过公众深思熟虑、审慎作出的决定”为前提,并补充说:“有许多提案显示出前景,包括联邦独立评估人员和监管机构、科技公司及其所创造产品的严格责任,以及一个独立机构来监管它们。”他们表示,联邦法律应确立最低保护标准,同时保留各州通过和执行更强措施的权利。

AI development should be conditioned on “deliberate, considered decision informed by the general public,” said the nonprofit, adding, “There are a number of proposals that show promise, including federal independent evaluators and regulators, strict liability for tech companies and the products they create, and an independent agency to regulate them.”Federal law, they stated, should establish minimum protections while preserving states’ authority to adopt and enforce stronger measures.

尽管这起诉讼是在加州法庭针对人工智能巨头提起的,但原告主张,暂停国内开发不会让科技竞争对手——尤其是中国——获胜。人工智能公司和美国总统唐纳德·特朗普一再表示,美国正全速推进,以在人工智能开发上击败北京——前提是华盛顿制定全球规则并启动与中国的谈判。

While the lawsuit targets AI titans in a California courtroom, the plaintiffs argue that pausing domestic development won’t hand a victory to tech rivals – notably China. AI firms and US President Donald Trump have repeatedly signalled the US is racing full throttle to beat Beijing in AI development — provided Washington sets global rules and kick-starts talks with China.

“我们预计,需要双边协议和一项国际条约,才能确保我们能够减少人工智能的负面影响并防止灾难发生,”不可替代教育基金会在给半岛电视台的一份声明中表示。

“We anticipate that bilateral deals and an international treaty will be needed to ensure that we can reduce the negative impacts of AI and prevent catastrophe,” the Irreplaceable Education Fund said in a statement to Al Jazeera.

“如果双方都以大致相同的幅度放缓,同时我们的人工智能失控风险降低,那么这并不会真正损害美国或中国的人工智能竞争力,反而会让我们所有人都更安全。”这一论点直面特朗普长期以来对放缓人工智能的抵制:他担心华盛顿方面表现出的任何单边克制都会让北京在全球人工智能发展中掌握主导权。

“If both sides are slowed by roughly the same factor, while the risk of losing control of our AIs is reduced, then this does not really harm the U.S. or China’s AI competitiveness, but it makes us all safer.”The argument confronts Trump’s longstanding resistance to slowing down AI: the fear that any unilateral restraint shown by Washington would allow Beijing to set the path in global AI development.

这位总统是人工智能快速扩张的支持者,此前曾表示,继续在人工智能演进中领先中国很重要,因为“谁赢得人工智能,谁就赢得一切”。

The president, a proponent of AI rapid expansion, has previously said that it was important to continue leading the evolution of AI over China, because “whoever wins AI wins”.

原告指出,尽管多年来不断有警告,国会迄今仍未通过监管人工智能的立法。

Despite years of warnings, Congress has yet to pass legislation governing AI, the plaintiffs noted.

“这些公司向超级智能人工智能狂奔的每一天,都是更多人和更多关键基础设施被置于风险之中的又一天,”这名原告说。现在,他们希望即将上任的民主党议员能听取他们的呼声。

“Every day these companies race forward towards super-intelligent AI is another day that more people and more critical infrastructure are put at risk,” said the plaintiff. Now, they’re hoping incoming Democratic lawmakers will heed their calls.

“新一届国会应当首先对新增人工智能开发实施冻结,”该组织表示。“然而,我们不能坐等议员们行动。法院必须提供救济。”

“The new Congress should start by implementing a freeze on new AI development,” the organisation said. “However, we can’t wait for lawmakers to act. Courts must provide relief.”

阅读媒体原文 ↗