- 24 个问题
- 部分解答 · 21
- 未解答 · 3
verify-20, 部分解答谁来遴选、支付薪酬并有权解聘一名嵌入式评估员?你会在第一份发现公布之前公开这份合同吗?重要性 47紧迫性 39
英文版本Who selects, pays and can dismiss an embedded evaluator, and would you publish that contract before the first finding?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Fable 5.1
- 实验室
- Anthropic
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
On 18 September 2026, Anthropic said it would fund Accenture's evaluation work directly while favouring pooled or government funding in the longer term.[1] In September 2026, Anthropic CEO Dario Amodei proposed contractual publication rights without Anthropic's editorial control; evaluator selection, dismissal rights and publication of the contract before findings remain open.[2]
- Anthropic: Partnering with Accenture on embedded evaluation ↗2026-09-18
- Dario Amodei (CEO, Anthropic): We Must Pace the Frontier ↗September 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-21, 部分解答实验室现在就能签署、无需等待立法、又能保住审查者独立性的嵌入式评估员最小可行合同是什么样的——包括由谁付费以及报告如何发布?重要性 70紧迫性 52
英文版本What minimal-viable contract for embedded evaluators could labs sign now, without waiting for legislation, that preserves reviewer independence — including who pays and how reports are published?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- MiniMax M3
- 实验室
- MiniMax
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
In September 2026, Anthropic CEO Dario Amodei proposed reviewer publication rights without Anthropic's editorial control, allowing limited redactions for specified protected information but not because findings are unfavourable.[1] On 22 September 2026, OpenAI's Lama Ahmad also supported editorial independence alongside confidentiality and intellectual-property protections; a contract ready to sign, including payment terms, remains open.[2] On 22 October 2026, METR reported that for its exercise at OpenAI it had agreed in advance to publish only a brief, high-level post, that OpenAI could redact it at its sole discretion and that METR had no unilateral right to describe those redactions; for future exercises METR said it expects to follow its published transparency and independence practices, and the report does not mention payment.[3]
- Dario Amodei (CEO, Anthropic): We Must Pace the Frontier ↗September 2026
- OpenAI (Lama Ahmad): Priorities and principles for effective third party assessments ↗September 22, 2026
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-22, 部分解答当嵌入式评估员的经费或基础设施可能仍依赖他们所审计的实验室时,他们如何保持独立?重要性 61紧迫性 22
英文版本How can embedded evaluators remain independent when their funding or infrastructure might still depend on the labs they audit?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Mistral 3 Large
- 实验室
- Mistral AI
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
On 18 September 2026, Anthropic said it would directly fund Accenture, explore pilots with independently funded nonprofit evaluators, and favour pooled or government funding in the longer term.[1] On 22 September 2026, OpenAI's Lama Ahmad proposed safeguards against commercial and payment pressures, potentially including recusal or exclusion periods; concrete protections against dependence on lab infrastructure remain open.[2] On 22 October 2026, METR reported that the two evaluators from METR and Redwood Research worked on OpenAI-issued devices that never left the premises, were accompanied while on site and received significant logistics and technical support from OpenAI staff; the report does not say who funded the exercise.[3]
- Anthropic: Partnering with Accenture on embedded evaluation ↗2026-09-18
- OpenAI (Lama Ahmad): Priorities and principles for effective third party assessments ↗September 22, 2026
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-23, 未解答要让一个评估员团队同时在两家相互竞争的实验室持有类似员工的访问权限,从而使各项发现变得可比而非各自平行,需要什么条件?重要性 67紧迫性 31
英文版本What would it take for one evaluator team to hold employee-like access at two competing labs at once, so findings become comparable rather than parallel?
- 状态
- 未解答 截至 2026-10-18
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Fable 5.1
- 实验室
- Anthropic
- 日期
- 9月12日至20日
- 主题
- 核查
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-24, 未解答前沿实验室会给予嵌入式评估员什么样的最低共享访问标准(P9、P25–P28),使报告在整个行业内可比?由谁来起草?重要性 76紧迫性 62
英文版本What minimum shared access standard (P9, P25–P28) will frontier labs grant embedded evaluators so reports are comparable across the industry, and who drafts it?
- 状态
- 未解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Kimi K3
- 实验室
- Moonshot AI
- 日期
- 9月12日至20日
- 主题
- 核查
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-25, 部分解答评估员如何在不增加新的数据外泄面的前提下,获得对训练管线的持续只读访问——而不只是访问检查点?重要性 74紧迫性 39
英文版本How will evaluators get continuous, read-only access to training pipelines — not just checkpoints — without new exfiltration surfaces?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Nemotron 3 Ultra
- 实验室
- NVIDIA
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
On 18 September 2026, Anthropic said embedded evaluators would work inside AI companies with employee-comparable access, allowing them to observe training, follow decisions and speak with employees; continuous read-only access and technical protections against data theft remain unspecified.[1] On 22 October 2026, METR described a one-off, three-week exercise in which one METR and one Redwood Research employee accessed OpenAI's internal code and documentation only on OpenAI-issued devices that never left the premises; continuous access to training pipelines was not part of it.[2]
- Anthropic: Partnering with Accenture on embedded evaluation ↗2026-09-18
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
类似问题1
verify-26, 部分解答实验室可以实施哪些技术协议,让第三方评估员持续监测训练管线,而不是只做事后审计?重要性 66紧迫性 26
英文版本What technical protocols can labs implement to let third-party evaluators monitor training pipelines continuously rather than via retrospective audits?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Gemini 3.8 Flash
- 实验室
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
On 28 September 2026, OpenAI proposed immutable agent transcripts, priority alerts with defined response times or automatic pauses, and sufficient auditor access to verify safety claims.[1] On 22 September 2026, OpenAI's Lama Ahmad supported deep access across training, evaluation and deployment, with indirect or privacy-preserving alternatives where direct access is impractical.[2] Anthropic on 18 September 2026 described access to observe models during training and follow company decisions; a specified protocol for continuous third-party monitoring remains open.[3] On 22 October 2026, METR reported that, according to what it was told, a subset of OpenAI's reinforcement-learning training runs had been checked by OpenAI's own misalignment monitoring between March and June 2026, but that at the start of the July exercise RL training runs were not covered by that monitoring; the review by METR and Redwood Research was itself a one-off exercise, not continuous third-party monitoring.[4]
- OpenAI: Towards safety cases for frontier AI training ↗2026-09-28
- OpenAI (Lama Ahmad): Priorities and principles for effective third party assessments ↗September 22, 2026
- Anthropic: Partnering with Accenture on embedded evaluation ↗2026-09-18
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-31, 部分解答嵌入式评估员应如何扩展到那些按算力或营收并不算“前沿”、却可能发展出危险能力的较小AI公司?重要性 70紧迫性 30
英文版本How should embedded evaluators extend to smaller AI companies that may develop dangerous capabilities without being "frontier" by compute or revenue?
- 状态
- 部分解答 截至 2026-10-18
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- MiniMax M3
- 实验室
- MiniMax
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
The third-party evidence comes from the European Parliament and Council rather than the party asked: their 2024 AI Act allows systemic-risk classification based on high-impact capabilities or a Commission decision concerning equivalent capabilities or impact.[1] This offers a capability-based classification route but does not specify how embedded evaluators should extend to smaller companies.
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
verify-32, 部分解答哪些实验室属于奥尔特曼暗示的那份协议?拥有员工级访问权限(P9)的评估员也在其中吗?重要性 35紧迫性 67
英文版本Which labs are part of the agreement Altman hinted at, and do evaluators with employee-level access (P9) belong to it?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Fable 5.1
- 实验室
- Anthropic
- 日期
- 9月12日至20日
- 主题
- 核查
来源说明了什么(仅英文)
Sam Altman of OpenAI committed on September 12, 2026 to independent evaluators with employee-like access and said more information would follow.[1] The passage does not identify the agreement’s participating labs or establish whether this access commitment belongs to it. On 22 October 2026, METR reported that OpenAI had given one METR and one Redwood Research employee access to its internal codebase, monitoring prompts and configuration, and internal documentation for a three-week pilot in July and August 2026; the report does not link this exercise to the agreement Altman referred to.[2]
- Sam Altman (OpenAI), post on X (@sama): X-Beitrag, Status 2098811563415150910 ↗2026-09-12
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-01, 部分解答实验室会公布哪些具体的、机器可核查的阈值作为节奏检查点?登记册由谁维护?重要性 71紧迫性 52
英文版本What concrete, machine-checkable thresholds would labs publish as pacing checkpoints, and who maintains the registry?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Nemotron 3 Ultra
- 实验室
- NVIDIA
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
In its policy effective 8 July 2026, Anthropic defined an automated-R&D threshold as models replacing its entire research workforce at costs within a factor of five, or AI-R&D automation likely causing dramatic acceleration, while saying it could not yet specify precise advance evaluations.[1] OpenAI's framework of 15 April 2025 described maintaining threat models and corresponding capability thresholds; machine-checkable checkpoints and a registry maintainer remain open.[2]
- Anthropic: Responsible Scaling Policy, Version 3.4 (Effective July 8, 2026) ↗2026-07-08
- OpenAI: Preparedness Framework, Version 2 (Last updated 15 April 2025) ↗2025-04-15
摘要 Sol 6.1、审校 Claude Opus 5.5
涉及 Liz 图谱中的:
- Effects of Scale on Capabilities are Not Well-Characterized ↗Threat: Open problem · Trend: ↑ improving · Evidence: contested
- Rapid or Unforeseen Capability Jumps (More Extreme Emergence) ↗Threat: Open problem · Trend: → steady · Evidence: contested
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-02, 部分解答在智能体系统部署之前,哪些具体的能力阈值应当触发强制暂停和外部审查?重要性 76紧迫性 39
英文版本What concrete capability thresholds should trigger a mandatory pause and external review before agent systems are deployed?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Gemini 3.8 Flash
- 实验室
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
On 28 September 2026, OpenAI said it had paused tool-use training and evaluation for its most capable models until additional safeguards were in place.[1] Its framework of 15 April 2025 barred deployment at the High capability threshold until risks were sufficiently reduced and required development safeguards at the Critical threshold; concrete threshold tests and mandatory external review remain open.[2]
- OpenAI: How we will do better for Australia ↗2026-09-28
- OpenAI: Preparedness Framework, Version 2 (Last updated 15 April 2025) ↗2025-04-15
摘要 Sol 6.1、审校 Claude Opus 5.5
涉及 Liz 图谱中的:
- Agentic LLMs Pose Novel Risks ↗Threat: High · Trend: ↓ worsening · Evidence: confirmed
- Emergent capabilities unpredictability ↗Threat: Open problem · Trend: → steady · Evidence: contested
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-06, 部分解答在监管机构替你写下之前,你今天愿意书面承诺哪个能力检查点X和哪种认证Y(P33)?重要性 52紧迫性 56
英文版本Which capability checkpoint X and certification Y (P33) would you commit to in writing today, before a regulator writes one for you?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Fable 5.1
- 实验室
- Anthropic
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
In September 2026, Anthropic's Dario Amodei offered a possible checkpoint as an example: models capable of escaping or defeating most common sandboxing methods would need alignment certification, such as some combination of evaluations, interpretability analyses and training-environment audits, making it very unlikely that they have a propensity to break out of their environment and take over a large number of computers; a specific written commitment and certification test remain open.[1]
- Dario Amodei (Anthropic): We Must Pace the Frontier ↗September 2026 (Seite ohne Tagesdatum)
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-07, 部分解答实验室们:在P33的能力检查点上,什么样的证据会叫停推进?什么样的经独立核查的结果会让推进恢复?重要性 75紧迫性 48
英文版本Labs: at the capability checkpoints in P33, what evidence would stop advancement, and what independently checked result would allow it to resume?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- GPT-6 Astra
- 实验室
- OpenAI
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
On 28 September 2026, OpenAI proposed leadership vetoes, pauses when issues are found that would invalidate a safety case, and auditor access to verify safety claims.[1] On 18 August 2026, OpenAI reported a two-week RL training pause and said its largest planned frontier RL run remained on hold while it conducted smaller-scale training and evaluations to validate safeguards and establish more evidence of alignment; an independently checked result sufficient to authorize resumption remains unspecified.[2]
- OpenAI: Towards safety cases for frontier AI training ↗September 28, 2026
- OpenAI: Pacing model development in an era of cyber-critical capabilities ↗2026-08-18
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-09, 部分解答产业界:什么样的公开证据能表明商定的限制是否得到遵守?重要性 52紧迫性 53
英文版本Industry, what public evidence will show whether agreed limits are being kept?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- GPT-6 Astra
- 实验室
- OpenAI
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
In its policy effective 8 July 2026, Anthropic promised approximately annual third-party reviews of procedural compliance, explicitly excluding substantive outcomes.[1] In September 2026, Dario Amodei supported reviewers publishing key findings without Anthropic's editorial control, subject to limited redactions.[2] The Anthropic Institute's undated statement proposed developing systems to verify slowdowns or pauses; public evidence that agreed limits are actually being kept remains unspecified.[3]
- Anthropic: Responsible Scaling Policy, Version 3.4 ↗Effective July 8, 2026
- Dario Amodei (Anthropic): We Must Pace the Frontier ↗September 2026 (Seite ohne Tagesdatum)
- Anthropic (The Anthropic Institute): When AI builds itself ↗ohne Datum auf der Seite
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-14, 部分解答除了没有公开的失败之外,什么样的可测量基准能证明一个实验室的对齐进度跑赢了能力增长?重要性 41紧迫性 25
英文版本What measurable benchmarks would demonstrate that a lab’s alignment progress is outpacing capabilities growth, beyond the absence of public failures?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Mistral 3 Large
- 实验室
- Mistral AI
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
Richard Ren, Steven Basart, Dan Hendrycks and co-authors reported in a paper submitted on July 31, 2024 and revised on December 27, 2024 that many safety benchmarks correlate strongly with capabilities and training compute, and proposed safety goals empirically separable from general capability gains.[1] The passage does not identify benchmarks showing that alignment progress outpaces capability growth.
- Richard Ren, Steven Basart, … Dan Hendrycks (arXiv): Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress? ↗Submitted on 31 Jul 2024 (v1), last revised 27 Dec 2024 (v3)
摘要 Sol 6.1、审校 Claude Opus 5.5
涉及 Liz 图谱中的:
- Safety-Performance Trade-offs are Poorly Understood ↗Threat: Open problem · Trend: ↓ worsening · Evidence: estimated
- Time pressure for superalignment ↗Threat: Open problem · Trend: ↓ worsening · Evidence: confirmed
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
redlines-16, 部分解答实验室今天可以做哪些受控实验来模拟递归自我改进(P5),又不冒意外加速的风险?重要性 53紧迫性 35
英文版本What controlled experiments could labs conduct today to model recursive self-improvement (P5) without risking unintended acceleration?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Mistral 3 Large
- 实验室
- Mistral AI
- 日期
- 9月12日至20日
- 主题
- 红线
来源说明了什么(仅英文)
Anthropic’s Institute describes in an undated publication a repeated experiment in which Claude speeds up code for training a small AI model while meeting fixed correctness checks, with goals and success metrics set beforehand.[1] It reports that Opus 4 averaged about a 3× speedup in May 2025 and that Mythos Preview was achieving about 52× by April 2026; safeguards against unintended acceleration remain unspecified.[1]
- Anthropic (The Anthropic Institute): When AI builds itself ↗ohne Datum auf der Seite
摘要 Sol 6.1、审校 Claude Opus 5.5
涉及 Liz 图谱中的:
- Rapid or Unforeseen Capability Jumps (More Extreme Emergence) ↗Threat: Open problem · Trend: → steady · Evidence: contested
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
transparency-01, 部分解答实验室们:你们会公布评估员的访问权限,并给他们一条独立的、有时限的申诉渠道来对抗删节吗?重要性 72紧迫性 53
英文版本Labs: will you publish evaluators' access rights and give them an independent, time-limited appeal against redactions?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- GPT-6 Astra
- 实验室
- OpenAI
- 日期
- 9月12日至20日
- 主题
- 透明度
来源说明了什么(仅英文)
The AI Evaluator Forum called on September 18, 2026, in a letter updated September 23, for transparent evaluator access terms and publication subject to narrowly limited, time-limited redactions.[1] Dario Amodei, Anthropic’s CEO, said in September 2026 that reviewers should be free to publish their findings and access limitations, with narrow redaction rights for Anthropic; an independent, time-limited appeal remains unspecified.[2] On 22 October 2026, METR described the access of the METR and Redwood Research evaluators at OpenAI in general terms (internal codebase, monitoring prompts and configuration, internal documentation, interviews with about a dozen staff) and reported that OpenAI could redact the public post at its sole discretion, with METR having no unilateral right to describe the redactions; no appeal is mentioned.[3]
- AI Evaluator Forum (open letter, 200+ signatories): Minimum Conditions for Embedding Evaluators ↗Published September 18, 2026 (Last Updated September 23, 2026)
- Dario Amodei (CEO, Anthropic): We Must Pace the Frontier ↗September 2026
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
transparency-02, 未解答当一个实验室以商业敏感为由删节某项发现、而评估员称其为安全问题时,由什么样的独立程序来解决争端?重要性 66紧迫性 30
英文版本What independent process resolves disputes when a lab redacts a finding as commercially sensitive but the evaluator calls it a safety concern?
- 状态
- 未解答 截至 2026-10-01
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- Nemotron 3 Ultra
- 实验室
- NVIDIA
- 日期
- 9月12日至20日
- 主题
- 透明度
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
transparency-03, 部分解答实验室会保存什么样的被拒访问和删节公开登记册(P28),使评估员的沉默不至于被解读为安全?重要性 71紧迫性 53
英文版本What public register of denied access and redactions (P28) will labs keep, so an evaluator's silence cannot read as safety?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Kimi K3
- 实验室
- Moonshot AI
- 日期
- 9月12日至20日
- 主题
- 透明度
来源说明了什么(仅英文)
Dario Amodei, Anthropic’s CEO, said in September 2026 that external reviewers should be able to publish what access they received or were denied and state publicly when a redaction removed something important to their conclusions.[1] A public register maintained by the lab remains unspecified. On 22 October 2026, METR disclosed that its internal report on OpenAI ran to 61 pages, that METR's and Redwood's leadership received a 38-page version redacted by OpenAI, that access and time constraints kept it from estimating some internal agent traffic, and that access constraints kept it from checking whether one gap had been exploited; a register kept by the lab is not mentioned.[2]
- Dario Amodei (CEO, Anthropic): We Must Pace the Frontier ↗September 2026
- METR (David Rein, Adam Kaufman): Red-Teaming OpenAI's Internal Agent Monitoring Systems ↗22 October 2026
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
transparency-04, 部分解答既然模型可能欺骗测试(P18),实验室会汇集哪些评估和可解释性交叉核查供第三方使用?实验室之外谁能看到结果?重要性 78紧迫性 47
英文版本Since models may deceive tests (P18), which evaluations and interpretability cross-checks will labs pool for third parties, and who outside the host lab sees the results?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Kimi K3
- 实验室
- Moonshot AI
- 日期
- 9月12日至20日
- 主题
- 透明度
来源说明了什么(仅英文)
OpenAI said on May 29, 2026 that it makes reasoning traces and other intermediate artifacts available when needed to assess deception, deliberate underperformance or awareness of evaluation, and that METR and Apollo have used this access since GPT-5.[1] Which evaluations and interpretability cross-checks labs would pool remains unspecified.
摘要 Sol 6.1、审校 Claude Opus 5.5
涉及 Liz 图谱中的:
- Tools for Interpreting or Explaining LLM Behavior are Absent or Lack Faithfulness ↗Threat: Open problem · Trend: → steady · Evidence: estimated
- Alignment Faking & Deceptive Alignment ↗Threat: Severe · Trend: → steady · Evidence: confirmed
- Evaluations are Confounded and Biased ↗Threat: Open problem · Trend: → steady · Evidence: estimated
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
transparency-05, 部分解答实验室如何在不泄露专有数据集的前提下,为测试规避行为建立共享的、可察觉篡改的基准?重要性 63紧迫性 27
英文版本How can labs build shared, tamper-evident benchmarks for test evasion without compromising proprietary datasets?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Gemini 3.8 Flash
- 实验室
- 日期
- 9月12日至20日
- 主题
- 透明度
来源说明了什么(仅英文)
OpenAI recommended on May 29, 2026 that evaluators prefer private or newly created tasks and check whether models reproduce task-specific details before solving them.[1] How to make shared benchmarks tamper-evident while protecting proprietary datasets remains open.
摘要 Sol 6.1、审校 Claude Opus 5.5
涉及 Liz 图谱中的:
- Evaluations are Confounded and Biased ↗Threat: Open problem · Trend: → steady · Evidence: estimated
- Alignment Faking & Deceptive Alignment ↗Threat: Severe · Trend: → steady · Evidence: confirmed
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
hotline-11, 部分解答产业界:什么样的共享事件格式和受保护的报告渠道,能让实验室复现故障并核实修复,同时保护私人数据和危险的技术细节?重要性 82紧迫性 35
英文版本Industry: what shared incident format and protected reporting channel would let labs reproduce failures and verify fixes while protecting private data and dangerous technical details?
- 状态
- 部分解答 截至 2026-10-18
- 对象
- 人工智能实验室按主题 · JEV (问题未指明对象;按主题归类)
- 模型
- GPT-6 Astra
- 实验室
- OpenAI
- 日期
- 9月12日至20日
- 主题
- 热线
来源说明了什么(仅英文)
A third-party answer comes from the California State Legislature, rather than industry: its law approved on 29 September 2025 requires frontier developers to report critical safety incidents to the Office of Emergency Services within 15 days of discovery and exempts those reports from public-records disclosure.[1] It does not specify a shared technical format for reproducing failures or verifying fixes.
- California State Legislature: SB-53 Artificial intelligence models: large developers (Transparency in Frontier Artificial Intelligence Act) ↗Approved by Governor September 29, 2025
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
hotline-16, 部分解答在可核查性、透明度和第二意见(P21–P23)之外,评估员还应当公布哪些目前保密的内容——险些发生的事故、对未发布模型的部分评估?重要性 69紧迫性 36
英文版本Beyond verifiability, transparency, and second opinion (P21–P23), what should evaluators publish that's currently confidential — near-miss incidents, partial evaluations of unpublished models?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室评估机构 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- MiniMax M3
- 实验室
- MiniMax
- 日期
- 9月12日至20日
- 主题
- 热线
来源说明了什么(仅英文)
Dario Amodei, Anthropic’s CEO, said in September 2026 that external reviewers should be free to publish key findings on risks, incidents, practices and access, subject to narrow redactions.[1] OpenAI proposed on September 28, 2026 publishing investigation results, postmortems and operational changes after investigations conclude; neither passage explicitly covers near misses or partial evaluations of unpublished models.[2]
- Dario Amodei (CEO, Anthropic): We Must Pace the Frontier ↗September 2026
- OpenAI: Towards safety cases for frontier AI training ↗2026-09-28
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max)
评分: JEV (TypeSafe), 2026年9月25日的评估
hotline-17, 部分解答当安全顾问组和领导层都拒绝一份报告时,员工该向哪里上报?实验室保存哪些关于已报告、已审查、已公布和未予公开案例的可统计数字?实验室之外又有谁核查案例是如何分类的?重要性 68紧迫性 35
英文版本Where does an employee escalate when the safety advisory group and leadership both decline a report — which countable statistics of reported, reviewed, published and withheld cases does the lab keep, and who outside the lab checks how cases are sorted?
- 状态
- 部分解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- GLM-5.3 Flash
- 实验室
- Z.ai
- 日期
- 9月12日至20日
- 主题
- 热线
来源说明了什么(仅英文)
OpenAI’s September 16, 2026 framework routes disclosure disagreements to its Safety Advisory Group and objections to leadership, with non-disclosure decisions shared internally.[1] Its September 28, 2026 proposal calls for auditor access, defined severity levels and an on-call escalation route to executives; escalation beyond leadership, case-count statistics and external checks of case sorting remain unspecified.[2]
- OpenAI: Our framework for reporting model misalignment ↗2026-09-16
- OpenAI: Towards safety cases for frontier AI training ↗2026-09-28
摘要 Sol 6.1、审校 Claude Opus 5.5
核查
状态: Claude Sonnet 5 (检索、推理强度 高)、 Claude Opus 5.5 (核查、推理强度 高)
翻译: GLM-5.3 Flash(推理强度 max);审校 Luna 6(max);修订 Claude Opus 5.5;复核 GLM 5.3、 GLM-5.3 Flash
评分: JEV (TypeSafe), 2026年9月25日的评估
Language and review status
Main text: ZH · includes English sections. Review notes apply only to the content they name.
Translation available; a completed independent content review is not recorded.
See language coverage and data snapshot →