Anthropic is letting METR researchers look under the hood with employee-level access

Anthropic is giving employee-level access to six researchers from METR, a nonprofit that tests frontier AI models, to independently assess Anthropic’s models, safety practices, and risk management, the company said in a blog post on Thursday.

Anthropic 在周四的一篇博客文章中表示，该公司已向非营利组织 METR 的六名研究人员授予了员工级别的访问权限。METR 致力于测试前沿人工智能模型，此次合作旨在让该组织能够独立评估 Anthropic 的模型、安全实践及风险管理。

The arrangement comes at a time where incidents of AI agents acting outside their instructions have been disclosed every week or two since late August. The incidents have sharpened calls for independent checks on AI companies.

这一安排出台之际，自 8 月下旬以来，AI 智能体违背指令行事的事件几乎每隔一两周就会被披露一次。这些事件加剧了外界对 AI 公司进行独立审查的呼声。

“We want to be publicly accountable for the decisions we’re making,” Anthropic said in the blog post. “We aim to close the gap between what evaluators understand about a frontier AI company’s systems and what the company itself understands.”

“我们希望为所做的决策承担公开责任，”Anthropic 在博客中写道，“我们的目标是缩小评估人员对前沿 AI 公司系统的理解与公司自身理解之间的差距。”

Under the new arrangement, the METR team has Anthropic-issued company laptops, office space, and access to internal tools roughly on par with what Anthropic’s own risk assessment teams use. “Anthropic employees are encouraged to share information with them just as they would with a fellow employee working on alignment risk” the company said.

根据这项新安排，METR 团队配备了 Anthropic 发放的公司笔记本电脑和办公空间，并拥有与 Anthropic 内部风险评估团队大致相当的内部工具访问权限。该公司表示：“我们鼓励 Anthropic 员工像对待从事对齐风险工作的同事一样，与他们分享信息。”

Anthropic compared the arrangement to banking, similar to how regulators place resident examiners inside the largest银行. METR will publish its own assessments of Anthropic’s risks, incidents, and practices, and the company has given up the right to block them.

Anthropic 将此安排比作银行业，类似于监管机构向大型银行派驻常驻审查员的做法。METR 将发布其对 Anthropic 风险、事件和实践的独立评估，而 Anthropic 已放弃了阻止这些评估发布的权利。

“Anthropic can’t veto publication, but we can require them to redact information if it falls into narrow, pre-defined categories,” the company said, such as customer data, anything attorney-client privileged, or relating to information about its model architectures.

“Anthropic 无权否决发布，但如果信息属于狭义的预定义类别，我们可以要求他们进行删减，”该公司表示，这些类别包括客户数据、受律师-客户特权保护的内容，或与模型架构相关的信息。

“METR is free to characterize our redactions publicly, and say whether it believes the redactions were justified, as long as it does so without revealing the redacted information” Anthropic added.

“METR 可以公开评价我们的删减行为，并说明其是否认为这些删减是合理的，前提是不能泄露被删减的具体信息，”Anthropic 补充道。

METR is the only evaluator so far at the top of a three-tier system Anthropic is building. Evaluators in the bottom tier, like biosecurity nonprofit SecureBio, run narrow reviews and require Anthropic’s approval to publish. The middle tier gets more information and can publish without approval, while the top tier adds METR’s employee-level access.

METR 是目前 Anthropic 正在构建的三级评估体系中，唯一处于最高级别的评估机构。底层评估机构（如生物安全非营利组织 SecureBio）仅进行有限的审查，且发布结果需经 Anthropic 批准。中层评估机构可获得更多信息并无需批准即可发布，而最高层级则增加了 METR 所享有的员工级访问权限。

Thursday’s announcement makes good on a promise Anthropic CEO Dario Amodei made in a September essay, where he proposed embedding outside evaluators inside AI labs as a first step toward slowing the technology’s development.

周四的公告兑现了 Anthropic 首席执行官 Dario Amodei 在 9 月份一篇文章中做出的承诺，他当时提议将外部评估人员嵌入 AI 实验室，作为减缓该技术发展的第一步。

OpenAI quickly pledged the same, and both companies, along with Google, Meta, Nvidia, and xAI, have since signed a White House accord committing to independent review.

OpenAI 迅速做出了同样的承诺。此后，这两家公司与谷歌、Meta、英伟达和 xAI 一起签署了白宫协议，承诺接受独立审查。