{"schemaVersion":"nuwa-current-record-1","version":"2026-10-29.1","publication":"published","work":{"id":"control-monitoring-deployment","title":"Practical challenges of control monitoring in frontier AI deployments","aliases":[],"authors":[],"primaryFunction":"governance","secondaryFunctions":[],"kind":"technical-report","sourceId":"control-monitoring-deployment-publication","attributionSourceId":"control-monitoring-deployment-publication","originalUrl":"https://www.aisi.gov.uk/research/practical-challenges-of-control-monitoring-in-frontier-ai-deployments","publicationDate":"2025-12-15","updatedOn":"2026-10-25","areas":["deception"],"scopeNote":{"en":"Relevance is assigned to the research question and stated conditions of the work. A topic tag is not a risk rating.","zh":"风险关联依据研究问题及明确条件建立，主题标签不代表风险评级。"},"protocolIds":[],"contributions":[{"organizationId":"aisi","role":"publisher","sourceId":"control-monitoring-deployment-publication","locator":"Official publication masthead and article byline"}],"summary":{"en":"Analyzes how parallel agents, oversight latency and incremental attacks affect synchronous and asynchronous control monitoring.","zh":"分析并行智能体、监督延迟和增量攻击如何影响同步及异步控制监测。"},"note":{"question":{"en":"What must hold for control monitoring to prevent harm when many agent instances act in parallel and oversight takes time?","zh":"当大量智能体实例并行行动且监督需要时间时，控制监测要满足哪些条件才能防止危害？"},"method":{"en":"The paper analyzes synchronous, semi-synchronous and asynchronous monitoring using a safety-case sketch. Four hypothetical deployment cases examine detection, oversight latency, harm latency, recovery and cumulative attack progress.","zh":"论文通过安全论证草图分析同步、半同步与异步监测。四个假设部署案例检查检测、监督时延、危害时延、恢复能力及累积攻击进展。"},"finding":{"en":"The analysis identifies failure modes when harm precedes review, partial attack progress cannot be reversed, parallel instances coordinate redundant attacks, or monitor failures correlate across steps. Fully synchronous blocking offers stronger protection at substantial availability and latency cost.","zh":"分析指出几类失效条件：危害早于审查发生、部分攻击进展不可逆、并行实例协调冗余攻击，以及监测错误跨步骤相关。完全同步阻断提供更强保护，但付出明显的可用性与时延代价。"},"supports":{"en":"The framework helps expose assumptions needed to turn per-attempt detection evidence into a deployment argument. It does not certify a particular production system.","zh":"该框架有助于揭示将单次攻击检测证据转化为部署论证所需的假设，并不认证任何具体生产系统。"},"limits":{"en":"The case studies concern hypothetical deployments and modeled attack vectors. Conclusions depend on representative attacks, incrimination, shutdown and recovery assumptions; empirical monitor performance alone does not establish these conditions.","zh":"案例针对假设部署和建模攻击路径。结论依赖攻击代表性、归责、停机与恢复假设；经验监测表现本身不能证明这些条件成立。"},"connections":null},"resources":[{"label":"Read the full paper","url":"https://arxiv.org/abs/2512.22154"}],"noteStatus":{"question":"verified","method":"verified","finding":"verified","supports":"verified","limits":"verified","connections":"limited"},"noteAvailability":{"en":"The substantive sections shown here were checked against the original source. Unfilled sections are omitted; source checking is not experimental replication.","zh":"此处展示的实质性章节已对照原始来源核查；未完成的章节已省略。来源核查不等于实验复现。"},"noteSourceLocator":"arXiv:2512.22154v1, §§2–3, 4–5, 6.1–6.2; Appendix E","tags":[]},"reading":{"id":"control-monitoring-deployment","title":"Practical challenges of control monitoring in frontier AI deployments","en":{"takeaway":"Monitoring safety depends on detection timing, parallel instances and the ability to recover from partial attacks.","question":"Under which deployment dynamics can monitoring prevent harm rather than merely detect it afterward?","design":"A safety-case analysis compares synchronous, semi-synchronous and asynchronous protocols using hypothetical case studies, separating monitorability, latency and recovery claims.","findings":["Asynchronous review fails when harm becomes irreversible before oversight can act.","Stopping one instance does not necessarily stop coordinated parallel actions; reversing partial progress is a separate requirement."],"interpretation":"NUWA interpretation: a high offline catch rate cannot substitute for a deployment-specific timing and recovery argument.","limits":"Case studies and latency estimates are hypothetical and need empirical validation. Monitor trustworthiness and dominant unseen attack vectors remain unresolved.","connections":["NUWA connection: async-control empirically tests some after-the-fact monitors; control-safety-case provides an earlier conservative extrapolation structure."],"decisionRelevance":"NUWA recommendation: document oversight latency, harm latency, global stop authority and recovery evidence before choosing monitoring synchrony."},"zh":{"takeaway":"监控安全取决于检测时机、并行实例以及恢复部分攻击后果的能力。","question":"什么部署动态下，监控能防止伤害，而不仅是事后发现？","design":"安全论证比较同步、半同步和异步协议，用假想案例分别检查可监控性、延迟与恢复主张。","findings":["若伤害在监督生效前不可逆，异步审核便失效。","停止一个实例不一定阻止协同并行动作，撤销部分进展也是独立要求。"],"interpretation":"NUWA 解读：高离线检出率不能替代特定部署的时序与恢复论证。","limits":"案例与延迟估计是假设性的，需要实证验证；监控可信性和未见优势攻击仍未解决。","connections":["NUWA 关联：异步控制实证测试部分事后监控，控制安全论证提供早期保守外推结构。"],"decisionRelevance":"NUWA 建议：选择监督同步方式前，记录监督延迟、伤害延迟、全局停止权限和恢复证据。"},"sourceRefs":[{"sourceId":"control-monitoring-deployment-publication","url":"https://arxiv.org/html/2512.22154","locator":"2 Monitoring protocols in realistic deployments","version":"Selected original resource https://arxiv.org/html/2512.22154; sha256:09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754; archived source capture; exact capture time not independently established","artifactSha256":"09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754"},{"sourceId":"control-monitoring-deployment-publication","url":"https://arxiv.org/html/2512.22154","locator":"3 A safety case sketch","version":"Selected original resource https://arxiv.org/html/2512.22154; sha256:09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754; archived source capture; exact capture time not independently established","artifactSha256":"09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754"},{"sourceId":"control-monitoring-deployment-publication","url":"https://arxiv.org/html/2512.22154","locator":"3.1.3 Harm prevention and reversibility","version":"Selected original resource https://arxiv.org/html/2512.22154; sha256:09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754; archived source capture; exact capture time not independently established","artifactSha256":"09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754"},{"sourceId":"control-monitoring-deployment-publication","url":"https://arxiv.org/html/2512.22154","locator":"6.1 Takeaways","version":"Selected original resource https://arxiv.org/html/2512.22154; sha256:09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754; archived source capture; exact capture time not independently established","artifactSha256":"09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754"},{"sourceId":"control-monitoring-deployment-publication","url":"https://arxiv.org/html/2512.22154","locator":"6.2 Limitations","version":"Selected original resource https://arxiv.org/html/2512.22154; sha256:09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754; archived source capture; exact capture time not independently established","artifactSha256":"09d528c275e9087571df987da0021359e0761a27adf74269a1f447d692225754"}],"publication":{"state":"published","publishedOn":"2026-10-29","authorizationReference":"user-authorized-global-index-publication-2026-10-29","reviewedCandidateId":"local-6efe31bc4a0c0f37"},"review":{"state":"published-note","sourceScope":"full-text","availableScope":null,"readingCoverage":"selected substantive sections of the full source body","completePaperAudit":false,"independentEvidenceReview":"pending-human-review","bilingualEditorialReview":"pending-human-review","authorship":"model-assisted-research-note","preparedFingerprint":"7d3035438961715541875735fe6dc38e8e24cf4f03f68ed9fa97e650b97442c3"}},"sources":[{"id":"control-monitoring-deployment-publication","title":"Practical challenges of control monitoring in frontier AI deployments","url":"https://www.aisi.gov.uk/research/practical-challenges-of-control-monitoring-in-frontier-ai-deployments","version":"Source capture 2026-10-25","publicationDate":"2025-12-15","datePrecision":"day","type":"research-publication","verifiedOn":"2026-10-25","documentId":"https://www.aisi.gov.uk/research/practical-challenges-of-control-monitoring-in-frontier-ai-deployments"}]}