风险治理 · 技术报告

A sketch of an AI control safety case

NUWA 研究解读 · 独立复核待完成

首次发表: 2025-01-28来源核验: 2026-10-25

原始来源 ↗引用本记录 ↓欺骗 →

NUWA Research Note

研究问题研究设计与假设来源发现 NUWA 解读局限与相反证据研究联系研究与决策意义机构与角色原始资源

模型辅助整理 · 选读原始全文或官方文章中的相关章节；尚未完成逐章节审计。

研究速读

该草图将控制主张分解为代理保真度、攻击强度及保守部署外推。

研究问题

即使内部受监控智能体试图外传敏感信息，哪些证据可支持它无法成功？

研究设计与假设

假想内部机器学习工程部署配对控制协议与红队替代模型，再将代理任务结果外推为部署论证。

来源发现

论证要求代理中的控制不强于部署，攻击不弱于可部署攻击。

并行协作、有状态或分散攻击是简单外推的明确反驳条件。

NUWA 解读

NUWA 解读：论证图组织待证明义务，本身不提供义务已满足的证据。

局限与相反证据

这是特定威胁和能力条件下的草图，不是完成的安全论证或部署实证认证。

研究联系

NUWA 关联：实际监控挑战细化同步与恢复；异步控制加入有界实证结果，但仍需部署论证。

研究与决策意义

NUWA 建议：逐项保守边界主张绑定版本化证据，称论证完成前列明反驳条件。

精确来源位置

A sketch of an AI control safety case · Executive SummarySelected original resource https://arxiv.org/html/2501.17315; sha256:e43991b6dedb523fbd63407bb1eca77baffa591fee85d8040960d66d0ed674f9; archived source capture; exact capture time not independently established
A sketch of an AI control safety case · 3 The safety caseSelected original resource https://arxiv.org/html/2501.17315; sha256:e43991b6dedb523fbd63407bb1eca77baffa591fee85d8040960d66d0ed674f9; archived source capture; exact capture time not independently established
A sketch of an AI control safety case · 6 Stages of a control evaluationSelected original resource https://arxiv.org/html/2501.17315; sha256:e43991b6dedb523fbd63407bb1eca77baffa591fee85d8040960d66d0ed674f9; archived source capture; exact capture time not independently established
A sketch of an AI control safety case · 7.2 Core control evaluation claimsSelected original resource https://arxiv.org/html/2501.17315; sha256:e43991b6dedb523fbd63407bb1eca77baffa591fee85d8040960d66d0ed674f9; archived source capture; exact capture time not independently established
A sketch of an AI control safety case · 8.1 Limitations of this safety case sketchSelected original resource https://arxiv.org/html/2501.17315; sha256:e43991b6dedb523fbd63407bb1eca77baffa591fee85d8040960d66d0ed674f9; archived source capture; exact capture time not independently established

来源核查、数据提取、编辑批准与实验复现分别记录。方法说明 →

机构与角色

研究机构英国 AI 安全研究所 Redwood Research

作者 (5)

Tomek Korbak · Joshua Clymer · Benjamin Hilton · Buck Shlegeris · Geoffrey Irving

引用与复用

引用研究发现时请引用原始出版物；引用 NUWA 的整理与解释时，单独引用本站记录及对应快照。

BibTeX ↓RIS ↓记录与来源元数据 · JSON ↓阅读覆盖 · JSON 提交更正

原始出版物 · BibTeX

@misc{control-safety-case,
  title = {A sketch of an AI control safety case},
  url = {https://www.aisi.gov.uk/research/a-sketch-of-an-ai-control-safety-case},
  author = {Tomek Korbak and Joshua Clymer and Benjamin Hilton and Buck Shlegeris and Geoffrey Irving},
  year = {2025},
  date = {2025-01-28},
  version = {Source capture 2026-10-25},
  note = {Source capture 2026-10-25}
}

NUWA 记录 · BibTeX

@misc{control-safety-case-nuwa,
  title = {A sketch of an AI control safety case — NUWA research record},
  url = {https://whitzard.tech/nuwa/whitzard-index/work/control-safety-case/},
  version = {2026-10-29.1},
  organization = {NUWA Frontier AI Safety Lab},
  howpublished = {Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--control-safety-case.json},
  note = {NUWA research note. Source-reading scope and independent review status are stated on the record. Snapshot 2026-10-29.1. Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--control-safety-case.json.}
}

证据快照 2026-10-29.1 · 索引更新 · RSS

参考文献

[1] A sketch of an AI control safety case. 2025-01-28 ; Executive Summary; 3 The safety case; 6 Stages of a control evaluation
[2] A sketch of an AI control safety case ; Author affiliations on the original paper title page