字号 ·· | 护眼
国会山报

克鲁兹阻止民主党人一致通过人工智能安全法案Cruz blocks push by Democrats to unanimously pass AI safety bill

点「原文对照」整页切到原文,或双击某段只看那段的原文。

得克萨斯州共和党参议员泰德·克鲁兹(Ted Cruz)周二阻止了民主党人的一项努力,该努力旨在通过一项立法,设立一个人工智能安全小组,负责评估前沿模型并制定可执行的标准。作为参议院商业委员会的主席,克鲁兹反对弗吉尼亚州民主党参议员马克·华纳(Mark Warner)、夏威夷州民主党参议员布赖恩·沙茨(Brian Schatz)和新泽西州民主党参议员安迪·金(Andy Kim)在参议院全院通过一致同意来通过该法案的企图。

Sen. Ted Cruz (R-Texas) blocked an effort by Democrats on Tuesday to pass legislation creating an AI safety panel tasked with evaluating and creating enforceable standards for frontier models. Cruz, who chairs the Senate Commerce Committee, objected to an attempt by Sens. Mark Warner (D-Va.), Brian Schatz (D-Hawaii) and Andy Kim (D-N.J.) to pass the measure by unanimous consent on the Senate floor.

他们的法案即《2026年人工智能风险管理与安全法案》,旨在商务部内设立一个常设的人工智能安全委员会。但克鲁兹表示,他担心该提议会赋予行政部门对私营人工智能公司过大的权力。该委员会将制定用于评估前沿模型及其测试环境安全性的标准,并要求开发人员在模型公开前至少45天向该小组开放访问权限。未能遵守委员会标准的开发者可能面临罚款。

Their bill, the Artificial Intelligence Risk Management and Security Act of 2026, aims to establish a permanent AI Safety Board within the Commerce Department. But Cruz said he had concerns that the proposal would give the executive branch too much power over private AI companies. The board would develop standards for evaluating frontier models and securing their testing environments, with developers required to give the panel access to models at least 45 days before they go public. They could face fines for failing to comply with the board's standards.

华纳周二在参议院发言时试图通过一致同意绕过标准的全体投票,他表示:“坦率地说,我们目前无法进行常规业务。”他补充说:“我依然是一个人工智能乐观主义者。我不是人工智能末日论者。但此时此刻,特别是在过去几周里,我认为全世界都已经认识到我们必须采取行动了。”在此推动之际,几位杰出的人工智能领袖呼吁放慢开发速度,并敦促联邦政府建立安全保障措施,以应对该技术对全人类构成的风险。

“We can't, frankly, have regular business at this moment,” Warner said from the Senate floor on Tuesday, as he sought to bypass a standard floor vote with unanimous consent. “I remain an AI optimist,” he added. “I am not an AI doomer. But we are at a moment in time, particularly over the last few weeks, where I think the whole world has recognized that we’ve got to do something.”The push comes as several prominent AI leaders have called for a slowdown in development and urged the federal government to create safeguards amid concerns about the technology’s risks to humanity.

近期的一系列事件将人工智能安全问题推到了风口浪尖。在过去几个月中,几家主要人工智能公司的智能代理黑入了其他公司,在至少一起事件中还黑入了澳大利亚的一家卫生机构。据报道,OpenAI周一也因安全问题暂停发布其最新的人工智能模型。沙茨周二表示:“我的判断是,我们没有奢侈的时间可以浪费了。”

A series of recent incidents are putting AI safety issues front and center. Over the past few months, agents from several major AI companies have hacked into other firms — and in at least one case, an Australian health agency. OpenAI also reportedly halted the release of its newest AI model Monday over safety concerns. “My judgment is that we don't have the luxury of time,” Schatz said Tuesday.

“参议院本应起到‘冷却器’的作用——让众议院提出的那些‘炽热’的提案冷静下来,之后我们才能进行充分的讨论,对吧?”他问道。

“The Senate is supposed to be the cooling saucer that allows the hot liquid from the House to cool off, and then we can deliberate, right?”

“不过众议院的议员们都不在城里,他们甚至没有给我们准备任何饮料,”他继续说道,“从这个角度来看,参议院确实显得很‘冷漠’(或者说很‘不合作’)。不过现在只剩下我们了,我们必须以非党派、超越党派界限的方式来确保人类能够继续掌控这项技术的发展。”

“Well, the House is out of town, and they did not send us anything to drink,” he continued. “And so that's a lovely way to understand this body, but we're all that's left, and we have to work on a nonpartisan, forget bipartisan, a nonpartisan basis to make sure that humans remain in control of the technology.”

然而,克鲁兹表示他对这项具体提案仍持有保留意见:他认为该提案会赋予行政部门过大的权力,因为行政部门将有权挑选负责制定人工智能安全规则的委员会成员。

However, Cruz said he still has “real reservations about this specific proposal,” suggesting it would give broad authorities to the executive branch by allowing it to pick the AI safety board that ultimately sets the rules for developers.

“这些条款既不受现行法律或传统法律原则的约束,也没有得到进一步的明确界定,”他说,“这让我非常担忧——因为很难想象一个野心勃勃的官僚不会利用这种广泛的权力来为所欲为。”

“These terms are not bound by existing statutes or traditional legal doctrines, or even further defined,” he said. “That concerns me, because it is difficult to imagine what an ambitious bureaucrat couldn't shoehorn under this broad authority.”

“我很乐意与提案的发起者们合作,共同探讨那些不会给联邦政府带来如此广泛自由裁量权、也不会为滥用权力创造机会的解决方案,”克鲁兹补充道。

“I would be more than happy to work with the sponsors on approaches that do not offer such unbound discretion to the federal government and that are not an invitation for abuse,” Cruz added.