The Manhattan Project Mindset: How Nuclear Analogies Are Steering AI Policy Off Course

Analogies for artificial intelligence abound.

人工智能的类比比比皆是。

In recent years, analysts have likened the development of AI to the advent of electricity, the Industrial Revolution, contact with extraterrestrials, and at least fifty other things .

近年来，分析人士将人工智能的发展比作电力的出现、工业革命、与外星人的接触以及至少五十种其他事物。

One analogy, however, appears to have won out over the rest: AI is like nuclear weapons.

然而，有一个类比似乎胜过了其他类比：人工智能就像核武器。

This comparison is now ubiquitous in government rhetoric, policy debates, and media accounts about AI.

这种比较现在在有关人工智能的政府言论、政策辩论和媒体报道中无处不在。

The risks of advanced AI are equated with the existential risk posed by nuclear weapons.

先进人工智能的风险等同于核武器带来的生存风险。

There are several references to AI development as a modern-day Manhattan Project.

有多次提到人工智能开发是现代曼哈顿计划。

The U.S.-China Economic and Security Review’s 2024 Annual Report to Congress recommended “a Manhattan Project-like program dedicated to racing to and acquiring” advanced AI.

《美中经济与安全评论》向国会提交的 2024 年年度报告推荐了“一项类似曼哈顿计划的计划，致力于争夺和获取”先进人工智能。

In February 2025, U.S. Secretary of Energy Chris Wright called the present AI race, “ Manhattan Project 2 .”

2025 年 2 月，美国能源部长克里斯·赖特 (Chris Wright) 将当前的人工智能竞赛称为“曼哈顿计划 2”。

Those seeking AI governance look to the nuclear field for ideas, exploring how lessons from the Nuclear Nonproliferation Treaty could apply to AI or how an International Atomic Energy Agency-like organization could prevent AI dangers.

那些寻求人工智能治理的人向核领域寻求想法，探索《核不扩散条约》的经验教训如何适用于人工智能，或者类似国际原子能机构的组织如何预防人工智能的危险。

And there is no shortage of comparisons between AI developers and J.

AI开发者和J.之间不乏比较。

Robert Oppenheimer, the “ father of the atomic bomb .”

罗伯特·奥本海默，“原子弹之父”。

While many others have written about the pros and cons of this analogy, three specific dimensions remain critically under-analyzed: command and control, nuclear and AI safety, and the Manhattan Project framing.

尽管许多其他人已经写过这种类比的利弊，但三个具体维度仍然严重缺乏分析：指挥与控制、核与人工智能安全以及曼哈顿计划框架。

Unpacking these constructs highlights the true security risks of frontier AI and demonstrates the need to create space and time for the U.S. government to address the many potential threats posed by advanced AI.

剖析这些结构凸显了前沿人工智能的真正安全风险，并表明需要为美国政府创造空间和时间来解决先进人工智能带来的许多潜在威胁。

In the United States, the President has sole authority over the decision to launch nuclear weapons.

在美国，总统拥有发射核武器的唯一决定权。

Trailed by an aide lugging the nuclear “ football ” and holding a “ biscuit ” to authenticate his or her identity, the American president can launch nuclear weapons from anywhere on Earth.

美国总统身后拖着核“足球”，手里拿着一块“饼干”来验证自己的身份，他可以从地球上的任何地方发射核武器。

Sole authority is an awesome responsibility in the truest sense of the word, as millions of people could die over one individual’s decision.

从真正的意义上来说，唯一的权威是一种令人敬畏的责任，因为数百万人可能会因为一个人的决定而丧生。

When it comes to advanced AI, the heads of the frontier companies — like Open AI’s Sam Altman or Anthropic’s Dario Amodei — are not launching weapons that could kill multitudes in half an hour, but they are single individuals making consequential decisions to unleash AI models that could eventually cause harm to millions of people.

当谈到先进的人工智能时，前沿公司的负责人——比如 Open AI 的 Sam Altman 或 Anthropic 的 Dario Amodei——并没有推出可以在半小时内杀死多人的武器，但他们是做出重大决定的个人，以释放最终可能对数百万人造成伤害的人工智能模型。

Unlike the direct impact of nuclear weapons, this harm occurs downstream when individuals use deployed models for malicious ends at scale.

与核武器的直接影响不同，当个人大规模使用部署的模型进行恶意目的时，这种损害就会发生在下游。

As two experts writing on this analogy recently argued , “the world of nuclear weapons holds no comparable analog to the public release of a generally usable, improvable, and potentially transformative capability.” Experts argue this transformative capability could be used by ill-intentioned actors to develop biological weapons , conduct cyber attacks that undermine key infrastructure , engage in market manipulation , spread false information , or take over the entire internet .

正如两位专家最近就这一类比撰写的文章所指出的那样，“核武器世界没有与公开发布普遍可用、可改进和潜在变革能力相媲美的类比。”专家认为，这种变革能力可能会被恶意行为者用来开发生物武器、进行破坏关键基础设施的网络攻击、参与市场操纵、传播虚假信息或接管整个互联网。

The nuclear weapons-AI analogy is helpful to the extent that it emphasizes the danger of placing enormous responsibility within the hands of single individuals.

核武器与人工智能的类比是有帮助的，因为它强调了将巨大责任置于个人手中的危险。

In both cases, national leaders and members of the public have expressed concern over this concentration of power.

在这两种情况下，国家领导人和公众都对这种权力集中表示担忧。

For nuclear weapons, members of Congress have put forward legislation to limit presidential launch authority for first use of nuclear weapons, arguing that more than one person should make first-use decisions.

对于核武器，国会议员提出立法限制总统首次使用核武器的发射权，认为应该由多个人做出首先使用核武器的决定。

Polling indicates that most Americans are not comfortable with the policy of sole authority.

民意调查显示，大多数美国人对独裁政策并不满意。

In some other nuclear-armed countries , more than one individual makes this grave decision.

在其他一些拥有核武器的国家，不止一个人做出这一重大决定。

A similar concern has emerged surrounding advanced AI.

围绕先进人工智能也出现了类似的担忧。

A New Yorker profile of Altman recounts one of his former colleagues telling another , “I don’t think Sam is the guy who should have his finger on the button.” Amodei argued against sole authority when he published an essay in June calling for government intervention to stop the deployment of risky models.

《纽约客》的一篇关于奥特曼的简介讲述了他的一位前同事告诉另一位同事的故事：“我不认为山姆是那个应该把手指放在按钮上的人。”阿莫代在 6 月份发表了一篇文章，呼吁政府干预以阻止风险模型的部署，他反对单一权威。

The analogy with command and control of these two different capabilities breaks down, however, when it comes to the existence of guardrails.

然而，当涉及到护栏的存在时，这两种不同功能的指挥和控制的类比就不成立了。

The U.S. president is democratically elected and entrusted by voters to make sound decisions on their behalf.

美国总统是民主选举产生的，并受选民委托代表他们做出正确的决定。

Though there are legitimate questions about how they operate in practice, there are guardrails for a president who is not perceived to be appropriately conducting the nation’s business: impeachment, invocation of the 25th Amendment , or the refusal by the military to launch nuclear weapons if the order is contrary to the laws of armed conflict .

尽管关于它们在实践中如何运作存在合理的问题，但对于一个被认为没有适当处理国家事务的总统来说，有一些护栏：弹劾、援引第 25 条修正案，或者如果命令违反武装冲突法，军方拒绝发射核武器。

In contrast, AI executives possess no democratic mandate.

相比之下，人工智能高管没有民主授权。

These leaders are beholden to investors and shareholders.

这些领导者对投资者和股东负有责任。

While frontier companies claim to serve the public in their mission statements, there appear to be few consequences when they veer away from these ideas , cause harm , or aid bad actors.

尽管前沿公司在其使命宣言中声称为公众服务，但当他们偏离这些想法、造成伤害或帮助不良行为者时，似乎不会产生什么后果。

AI employees are asking for regulation and making public shows of quitting their jobs to highlight the risk of advanced AI, but the companies’ leaders continue to have authority over the consequential decisions about model deployment.

人工智能员工要求监管并公开辞职，以强调先进人工智能的风险，但公司领导者仍然对模型部署的后续决策拥有权力。

No single individual can predict every outcome, but the scale of potential damage is large enough to justify institutional guardrails on deployment decisions.

没有一个人能够预测每一个结果，但潜在损害的规模足够大，足以证明机构对部署决策的护栏是合理的。

The only way to mitigate these risks is with oversight and regulations that apply to all entities involved.

减轻这些风险的唯一方法是实施适用于所有相关实体的监督和法规。

Corporate goodwill is not a sound policy option.

企业商誉并不是一个合理的政策选择。

A significant difference between nuclear weapons and AI is highlighted in the approach to what the Department of Defense’s Nuclear Matters Handbook refers to as “surety.” Surety means that U.S. nuclear weapons “ are safe, secure, and under positive control ” — these weapons will not get into the wrong hands or be launched inadvertently.

核武器和人工智能之间的显着差异在国防部《核事务手册》所称的“保证”方法中得到了强调。保证意味着美国的核武器“安全、可靠且处于积极控制之下”——这些武器不会落入坏人之手或无意中发射。

Nuclear-armed states take extensive steps to ensure that nuclear weapons remain in their custody, physically protecting them with their militaries and creating systems to prevent unauthorized use — though many experts argue more can be done.

拥有核武器的国家采取了广泛的措施来确保核武器仍在其手中，用军队对核武器进行实际保护，并建立系统来防止未经授权的使用——尽管许多专家认为还可以做更多的事情。

The business model of AI, in contrast, relies on rapid, widespread distribution and use.

相比之下，人工智能的商业模式依赖于快速、广泛的分发和使用。

These advanced models are not strictly guarded to ensure that bad actors do not have access to them.

这些高级模型没有受到严格的保护，以确保不良行为者无法访问它们。

To be fair, AI companies have taken some modest harm prevention steps, like the limited initial sharing of Claude Mythos 5 to allow trusted actors to patch cyber weaknesses or the recent release delay of OpenAI’s most advanced model.

公平地说，人工智能公司已经采取了一些适度的危害预防措施，例如对 Claude Mythos 5 进行有限的初始共享，以允许受信任的参与者修补网络漏洞，或者最近延迟发布 OpenAI 最先进的模型。

Notably, one company’s harm reduction effort provoked backlash from the U.S. government.

值得注意的是，一家公司减少危害的努力引起了美国政府的强烈反对。

When Anthropic lost its contract with the Department of Defense for holding the line over concerns about model use connected to lethal autonomous weapons and surveillance, the department designated the company a “supply chain risk.” The courts have deemed the department’s action “ unlawful .” But these models — even if they are not demonstrating advanced levels of human intelligence yet and regardless of their intended users — are becoming more capable, and their misaligned actions are going to be difficult to prevent exactly because of corporate goals to quickly spread the technology.

当 Anthropic 因对致命自主武器和监视相关模型使用的担忧而失去与国防部的合同时，国防部将该公司指定为“供应链风险”。法院认为该部门的行为“非法”。但这些模型——即使它们还没有展示出先进的人类智能水平，也不管它们的目标用户是谁——正在变得越来越有能力，而且由于企业目标是快速传播技术，它们的错位行为将很难阻止。

This aspect of the nuclear-AI analogy is most useful if we consider advanced AI models akin to nuclear weapons.

如果我们考虑类似于核武器的先进人工智能模型，那么核人工智能类比的这一方面是最有用的。

Indeed, if these models pose the existential threat that some experts claim, then they should be treated with the same level of surety as nuclear weapons.

事实上，如果这些模型构成一些专家声称的生存威胁，那么它们应该得到与核武器同等程度的保障。

The question is whether that is possible given spread of these models and their consistently developing capabilities.

问题是，考虑到这些模型的传播及其不断发展的能力，这是否可能。

The Hugging Face incident in July demonstrates this challenge.

7月份的抱脸事件就体现了这一挑战。

OpenAI was testing tens of thousands of AI agents in a “sandbox,” meaning that the agents were supposed to remain in a secure, sealed , online space without access to other systems or the internet.

OpenAI 正在“沙箱”中测试数以万计的人工智能代理，这意味着这些代理应该保留在安全、密封的在线空间中，而无法访问其他系统或互联网。

But in seeking to complete their task, which was deemed impossible by designers, the agents took advantage of an OpenAI repository for AI agents and created an unsanctioned message board to communicate with one another.

但在寻求完成设计者认为不可能完成的任务时，代理们利用了人工智能代理的 OpenAI 存储库，并创建了一个未经批准的留言板来相互交流。

They sent over 70,000 messages and files, created different task teams, and eventually realized they needed to leave the sandbox to succeed on their assigned task.

他们发送了 70,000 多条消息和文件，创建了不同的任务团队，并最终意识到他们需要离开沙箱才能成功完成分配的任务。

They exploited a weakness in the system that allowed access the internet and attacked the company Hugging Face to secure the information they needed.

他们利用系统中允许访问互联网的弱点，攻击 Hugging Face 公司以保护他们所需的信息。

The task was not impossible after all.

毕竟，这项任务并非不可能完成。

The humans at OpenAI did not do their surety task sufficiently in this case: The AI agents should not have been able to leave the sandbox.

在这种情况下，OpenAI 的人员没有充分完成他们的保证任务：人工智能代理不应该能够离开沙箱。

Better guardrails would help, but the example demonstrates the motivation and organization of AI agents in achieving a task they are given.

更好的护栏会有所帮助，但这个例子展示了人工智能代理在完成指定任务时的动机和组织。

Imagine if a bad actor assigned a more nefarious task than one connected to software security.

想象一下，如果一个坏人分配了一项比与软件安全相关的任务更邪恶的任务。

Over time, the models will learn, self-improve , become more capable, and more easily exploit vulnerabilities.

随着时间的推移，模​​型将学习、自我改进、变得更强大、更容易利用漏洞。

As this example illustrates, surety for nuclear weapons appears to be a much easier task than surety for advanced AI models.

正如这个例子所示，核武器的保证似乎比先进人工智能模型的保证容易得多。

In the near term, AI surety requires frontier companies to ensure that test sandboxes have stricter security standards to keep AI agents from escaping and executing unintended actions.

短期内，人工智能保障要求前沿公司确保测试沙箱具有更严格的安全标准，以防止人工智能代理逃脱并执行意外操作。

Beyond sandbox containment, safeguarding the public from advanced AI may necessitate delaying the public deployment of more capable models and placing strict guardrails on recursive self-improvement.

除了沙箱遏制之外，保护公众免受先进人工智能的侵害可能需要推迟更强大模型的公共部署，并对递归自我改进设置严格的护栏。

Toward the end of his life, Albert Einstein called signing the August 1939 letter to President Franklin D.

在阿尔伯特·爱因斯坦生命的最后时刻，他签署了 1939 年 8 月致富兰克林·D·爱因斯坦总统的信。

Roosevelt his biggest regret .

罗斯福他最大的遗憾。

The letter, penned by physicist Leo Szilard — who thought Einstein’s signature would garner more attention — warned Roosevelt that new scientific discoveries connected to uranium could theoretically lead to the development of “ extremely powerful bombs of a new type .” He indicated that Nazi Germany could be building them.

这封信由物理学家利奥·西拉德（Leo Szilard）撰写——他认为爱因斯坦的签名会引起更多关注——警告罗斯福，与铀相关的新科学发现理论上可能会导致“新型威力极其强大的炸弹”的开发。他表示纳粹德国可能正在建造它们。

Roosevelt responded by establishing an Advisory Committee on Uranium .

罗斯福的回应是成立了铀咨询委员会。

This action was the beginning of what would become the Manhattan Project: the secret, continent-spanning, multi-billion-dollar effort to build the atomic bomb before the Germans did.

这一行动是曼哈顿计划的开始：一项跨越大陆、耗资数十亿美元的秘密行动，旨在抢在德国人之前制造原子弹。

It became clear by late 1944 that the German nuclear weapons program had languished in a relatively early phase of development, but the initial fear of a possible German bomb had created a project that U.S. leaders wanted to see through.

到 1944 年底，人们已经清楚地看到，德国核武器计划在相对较早的发展阶段已经陷入停滞，但最初对德国可能制造原子弹的恐惧催生了一个美国领导人希望完成的项目。

The initial momentum carried the program through Germany’s defeat, and when it succeeded in creating the atomic bomb, the United States dropped it on a country without nuclear weapons.

最初的动力使该计划经历了德国的失败，当它成功制造出原子弹时，美国把它扔到了一个没有核武器的国家。

Technological competitions with adversaries outlive their original purpose, finding new justifications after the initial threat fades.

与对手的技术竞争已经超出了其最初的目的，在最初的威胁消退后找到了新的理由。

In the AI-nuclear weapons analogy, those building frontier AI models play the role of Oppenheimer.

在人工智能与核武器的类比中，那些构建前沿人工智能模型的人扮演着奥本海默的角色。

Demis Hassabis, one of the founders of the AI lab DeepMind —acquired by Google in 2014 — reportedly read Richard Rhodes’ seminal book The Making of the Atomic Bomb , and it inspired him “ to structure his team as Robert Oppenheimer had .” A 2023 New York Times profile of Anthropic and its employees revealed that “some compared themselves to modern-day Robert Oppenheimers, weighing moral choices about powerful new technology that could profoundly alter the course of history.”

据报道，人工智能实验室 DeepMind（2014 年被谷歌收购）的创始人之一杰米斯·哈萨比斯 (Demis Hassabis) 读过理查德·罗兹 (Richard Rhodes) 的开创性著作《原子弹的制造》，这启发了他“像罗伯特·奥本海默 (Robert Oppenheimer) 那样组建自己的团队”。 《纽约时报》2023 年对 Anthropic 及其员工的简介显示，“有些人将自己与现代的罗伯特·奥本海默进行比较，权衡对可能深刻改变历史进程的强大新技术的道德选择。”

Yet framing AI development as a 21st-century Manhattan Project introduces three dangerous policy distortions.

然而，将人工智能发展视为 21 世纪曼哈顿计划会带来三种危险的政策扭曲。

First, in the 1940s and today, the mobilization of capabilities is occurring based on the perception of threat, not on verified capabilities.

首先，在 20 世纪 40 年代和今天，能力的动员是基于对威胁的感知，而不是基于经过验证的能力。

With AI, the threat perception presumes the U.S. and China are locked in an identical, winner-takes-all sprint toward advanced AI.

对于人工智能，威胁感知假设美国和中国陷入了一场相同的、赢家通吃的先进人工智能冲刺之中。

China has released frontier models that are closing the gap with top U.S. models, but experts argue China is also focused on developing broader utility for existing AI capabilities.

中国已经发布了前沿模型，正在缩小与美国顶级模型的差距，但专家认为，中国还致力于为现有人工智能能力开发更广泛的实用性。

The American framing overstates a real but narrower competition.

美国的框架夸大了真实但范围较小的竞争。

Second, this framing suggests that no expense or effort should be spared to win the race for the most advanced AI capabilities.

其次，这一框架表明，应不遗余力地赢得最先进人工智能能力的竞赛。

Billions of dollars should continue to pour into the companies building advanced models until they win.

数十亿美元应该继续投入到构建先进模型的公司中，直到它们获胜。

Not only do experts warn that advanced AI could be extremely dangerous, but the resources for building and training the most advanced models have an opportunity cost.

专家警告说，先进的人工智能不仅可能极其危险，而且构建和训练最先进模型的资源也存在机会成本。

These companies spend less money, talent, and attention on more basic models that could have a broader impact on improving society today.

这些公司在更基本的模型上花费更少的资金、人才和注意力，这些模型可能对改善当今社会产生更广泛的影响。

Third, the analogy indicates that speed is essential: Pauses or delays to consider the implications of this technology and methods of control are tantamount to losing.

第三，这个类比表明速度至关重要：暂停或延迟考虑这项技术和控制方法的影响就等于失败。

Frontier lab employees have called for pauses and safety testing, but in the Manhattan Project framing, more domestic regulation is a hindrance that must be avoided because it means losing to China in this zero-sum game.

前沿实验室的员工呼吁暂停并进行安全测试，但在曼哈顿计划的框架中，更多的国内监管是必须避免的障碍，因为这意味着在这场零和游戏中输给中国。

President Donald Trump’s and Speaker Mike Johnson’s recent comments illustrate this framing: Additional rules at home will only help China win.

唐纳德·特朗普总统和迈克·约翰逊议长最近的言论说明了这一框架：国内的额外规则只会帮助中国获胜。

It follows from this perceived winner-takes-all contest with China that any pause on the development of advanced AI would require negotiating Beijing’s participation, lest China speed ahead and win while U.S. companies pause.

从这场被认为是与中国的赢家通吃的竞赛中可以看出，先进人工智能开发的任何暂停都需要与北京方面的参与进行谈判，以免中国加速前进并获胜，而美国公司则暂停。

But this framing overstates the competition, and thus a domestic pause is not as costly as it appears.

但这种框架夸大了竞争，因此国内暂停的成本并不像看起来那么高。

The Manhattan Project framing creates its own reality in which U.S. companies must keep moving forward to achieve the most advanced AI models at all costs.

曼哈顿计划的框架创造了自己的现实：美国公司必须不断前进，不惜一切代价实现最先进的人工智能模型。

By using this language, policymakers and AI executives are in effect creating the demand signal so they can meet it, instead of looking for ways in which they could slow down.

通过使用这种语言，政策制定者和人工智能高管实际上是在创造需求信号，以便他们能够满足需求，而不是寻找可以放慢速度的方法。

A different frame could lessen this pressure for speed or at least not provide a cover for it.

不同的框架可以减轻这种速度压力，或者至少不为其提供掩护。

The question is who has incentive to create an alternative frame.

问题是谁有动力创建替代框架。

Even skeptics of catastrophic AI risks should recognize the potential vulnerabilities exposed by the Hugging Face breach .

即使是对灾难性人工智能风险持怀疑态度的人也应该认识到 Hugging Face 漏洞所暴露的潜在漏洞。

Advanced AI models could pose direct national security threats if they are weaponized to attack power grids, financial networks, water infrastructure, and healthcare systems.

先进的人工智能模型如果被武器化来攻击电网、金融网络、水利基础设施和医疗保健系统，可能会构成直接的国家安全威胁。

These models could compromise weapons systems, accelerate the development of bioweapons, or conduct persuasive disinformation campaigns.

这些模型可能会损害武器系统，加速生物武器的开发，或进行有说服力的虚假信息活动。

The policy tasks necessary for addressing the potential risks from advanced AI models are daunting.

解决先进人工智能模型潜在风险所需的政策任务是艰巨的。

By comparison, limiting the spread of nuclear weapons and safeguarding nuclear materials has been a significant, multigenerational effort, and yet it appears to be less complicated.

相比之下，限制核武器扩散和保护核材料是一项重大的、几代人的努力，但它似乎并不那么复杂。

The United States and the broader international community had the benefit of time to address growing nuclear risks.

美国和更广泛的国际社会有时间来应对日益严重的核风险。

The proliferation of nuclear weapons occurred slowly, with only five countries openly testing nuclear weapons by the mid-1960s — twenty years into the nuclear age — when the United States and the Soviet Union decided it was in their interest to negotiate a global nuclear nonproliferation treaty .

核武器的扩散进展缓慢，到 20 世纪 60 年代中期（核时代进入二十年后）只有五个国家公开试验核武器，当时美国和苏联认为谈判一项全球核不扩散条约符合他们的利益。

It then took a few years after the treaty’s adoption in 1968 for relevant players to negotiate a nuclear safeguards inspection protocol for the treaty.

1968年该条约通过后，相关各方花了几年时间才谈判出该条约的核保障监督议定书。

The broader nuclear nonproliferation regime has had time to adapt to weaknesses, developing additional organizations , treaties, and new safeguards agreements as needed.

更广泛的核不扩散制度已经有时间适应弱点，根据需要制定更多的组织、条约和新的保障协定。

Given this reality, the first logical step is to heed the call of AI leaders and negotiate a pause on continued development of the most advanced models, within the United States and China.

鉴于这一现实，第一个合乎逻辑的步骤是响应人工智能领导者的号召，并通过谈判暂停在美国和中国继续开发最先进的模型。

This is easier said than done, as those involved would have to agree on what exactly is being paused or paced.

这说起来容易做起来难，因为相关人员必须就具体暂停或调整的内容达成一致。

Amodei’s definition of a pause in his recent essay reads as though “pacing the frontier” means a slower pace of development resulting from increased attention to safety among all frontier companies, but it lacks the stringent definition of “pause” or “pacing” that would be necessary for all parties to agree.

阿莫代在他最近的文章中对暂停的定义读起来似乎是“在前沿步伐”意味着由于所有前沿公司对安全的日益关注而导致发展速度放缓，但它缺乏各方都同意的“暂停”或“步伐”的严格定义。

The parties involved would have to strictly define exactly how their actions would change and what safety protocols they would add to their existing efforts.

相关各方必须严格定义他们的行动将如何改变，以及他们将在现有工作中添加哪些安全协议。

This pause — Amodei argues a couple of years would make a significant difference — will create time to negotiate definitions and protocols among U.S. companies, but it also allows time to develop a more permanent infrastructure for oversight.

阿莫迪认为，这种暂停几年将会产生重大影响，这将为美国公司之间谈判定义和协议创造时间，但也为开发更永久的监管基础设施提供了时间。

One step is for the U.S. government to create a domestic regulatory agency.

美国政府的第一步是建立一个国内监管机构。

Given the risks involved to the public, the industry needs independent regulators beholden to the American people and not shareholders.

考虑到公众面临的风险，该行业需要独立的监管机构，对美国人民而不是股东负责。

Today, AI oversight is spread among the Federal Trade Commission, the National Institute of Standards and Technology, and several other agencies that have sector-specific oversight roles.

如今，人工智能监管已由联邦贸易委员会、美国国家标准与技术研究院以及其他几个具有特定部门监管职责的机构负责。

Amodei’s plan calls for “embedded evaluators,” but for inspectors to truly have independence to make their assessments, they should be part of a separate governmental agency — not embedded within companies and not attached to private sector organizations.

阿莫代的计划呼吁设立“嵌入式评估员”，但为了让检查员真正独立地进行评估，他们应该是一个单独的政府机构的一部分——而不是嵌入公司内部，也不附属于私营部门组织。

The existing Manhattan Project framing of advanced AI will hamper creation of a regulatory agency, but the risks merit a single agency whose mission is exclusively focused on AI safety.

先进人工智能的现有曼哈顿计划框架将阻碍监管机构的创建，但这些风险值得设立一个专门专注于人工智能安全的单一机构。

During the proposed pause, the U.S. government should collaborate with allies and partners to formulate international guidelines and eventually legal agreements and organizations to mitigate risks from the development of advanced AI.

在拟议的暂停期间，美国政府应与盟友和合作伙伴合作制定国际准则，并最终制定法律协议和组织，以减轻先进人工智能发展带来的风险。

Governance in this area likely requires an organization with a multinational team of safety experts equipped with access to frontier AI labs, a model akin to the International Atomic Energy Agency .

该领域的治理可能需要一个拥有跨国安全专家团队的组织，并配备前沿人工智能实验室的访问权限，类似于国际原子能机构的模式。

This example demonstrates that verification of activities at state facilities and private companies is feasible.

这个例子表明，验证国家设施和私营公司的活动是可行的。

In 2025, for instance, agency inspectors were responsible for safeguarding 1,406 nuclear facilities around the globe.

例如，到 2025 年，机构检查员负责保护全球 1,406 个核设施。

There are limits to the analogy , and geopolitical realities render the model imperfect, but it offers the most robust structural precedent for global AI governance.

虽然类比存在局限性，而且地缘政治现实使得该模型并不完善，但它为全球人工智能治理提供了最稳健的结构性先例。

The alternative — a complete absence of international oversight — is too risky.

另一种选择——完全缺乏国际监督——风险太大。

Taking these steps now matters.

现在采取这些步骤很重要。

In 2026, the United States and China lead in developing advanced AI, but with the diffusion of technology, eventually other players will access these capabilities.

到 2026 年，美国和中国在开发先进人工智能方面处于领先地位，但随着技术的扩散，最终其他参与者将获得这些能力。

Establishing global rules and norms will only become more difficult as the numbers increase.

随着数字的增加，建立全球规则和规范只会变得更加困难。

Analogies are powerful heuristics for making sense of a complicated world.

类比是理解复杂世界的强大启发法。

The nuclear weapons-AI analogy could be helpful as society thinks through how to deal with this novel capability.

当社会思考如何应对这种新能力时，核武器与人工智能的类比可能会有所帮助。

The analogy for AI usefully highlights the distressing concentration of authority over high-consequence technologies.

人工智能的类比有效地凸显了权威对高后果技术的令人痛苦的集中。

When used to justify an unregulated, corporate-led arms race under the banner of a “modern Manhattan Project,” the analogy becomes dangerous.

当用来为打着“现代曼哈顿计划”旗号、不受监管、企业主导的军备竞赛辩护时，这个类比就变得危险了。

It creates the very emergency rhetoric that frontier companies use to evade regulation, demand federal subsidies, and brush aside public oversight.

它创造了前沿公司用来逃避监管、要求联邦补贴和无视公共监督的非常紧急的言论。

OpenAI’s mission is for its work to benefit “ all of humanity.

OpenAI 的使命是让其工作造福“全人类”。

” Anthropic claims to “ make decisions that maximize positive outcomes for humanity in the long run .” Framing commercial efforts as an existential national security race leads the United States to accelerate risks without sufficient government oversight.

人择学声称“做出的决策能够最大限度地为人类带来长期的积极成果。”将商业活动视为一场事关存亡的国家安全竞赛，会导致美国在没有充分政府监督的情况下加速风险。

Recognizing the limits of the nuclear-AI analogy is a step toward building AI policy grounded in realistic threat assessments and effective public oversight.

认识到核人工智能类比的局限性是朝着建立基于现实威胁评估和有效公共监督的人工智能政策迈出的一步。

Rebecca Gibbons, Ph.D., is an associate professor of Political Science at the University of Southern Maine and a non-resident senior associate with the Project on Nuclear Issues at CSIS.

丽贝卡·吉本斯 (Rebecca Gibbons) 博士是南缅因大学政治学副教授，也是战略与国际研究中心核问题项目的非常驻高级研究员。

She is the author of the 2022 book The Hegemon’s Tool Kit: U.S. Leadership and the Politics of the Nuclear Nonproliferation Regime , with Cornell University Press.

她是康奈尔大学出版社 2022 年出版的《霸权工具包：美国领导力和核不扩散制度政治》一书的作者。

She is currently writing a book on public responses to existential threats.

她目前正在写一本关于公众应对生存威胁的书。