{"entities":[{"id":"openai","name":"OpenAI"},{"id":"anthropic","name":"Anthropic"},{"id":"google-deepmind","name":"Google DeepMind"},{"id":"xai","name":"xAI"},{"id":"amazon","name":"Amazon"},{"id":"eliezer-yudkowsky","name":"Eliezer Yudkowsky"},{"id":"fsf-2024","name":"Frontier Safety Framework (2024)"},{"id":"google","name":"Google"},{"id":"rsp-v2-0","name":"Anthropic Responsible Scaling Policy — v2.0"},{"id":"rsp-v2-1","name":"Anthropic Responsible Scaling Policy — v2.1"},{"id":"rsp-v2-2","name":"Anthropic Responsible Scaling Policy — v2.2"},{"id":"rsp-v3-1","name":"Anthropic Responsible Scaling Policy — v3.1"},{"id":"fsf-v2-0","name":"Google DeepMind Frontier Safety Framework — v2.0"},{"id":"fsf-v3-1","name":"Google DeepMind Frontier Safety Framework — v3.1"},{"id":"stargate-abilene","name":"Stargate I, Abilene"},{"id":"jan-leike","name":"Jan Leike"},{"id":"reward-modeling-paper","name":"Scalable agent alignment via reward modeling: a research direction"},{"id":"openai-lp-announcement","name":"OpenAI LP — capped-profit structure announcement (2019)"},{"id":"spacex","name":"SpaceX"},{"id":"meta","name":"Meta"},{"id":"alibaba-group","name":"Alibaba Group / 阿里巴巴集团"},{"id":"jingren-zhou","name":"Jingren Zhou"},{"id":"ms-fgf-2025","name":"Microsoft Frontier Governance Framework — Version 1 (February 2025)"},{"id":"openai-inc","name":"OpenAI, Inc. (OpenAI Foundation)"},{"id":"inflection-ai","name":"Inflection AI"},{"id":"extropians-1996-09-03-bostrom","name":"Re: Thinking about the future..."},{"id":"agi-ruin-2022","name":"AGI Ruin: A List of Lethalities"},{"id":"strategic-concealment","name":"Strategic concealment"},{"id":"harm-requires-hostility","name":"Without hatred it will not harm us"},{"id":"experiment-and-see","name":"Run experiments and see what happens"},{"id":"extropians-2000-09-05-yudkowsky-4292","name":"Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...)"},{"id":"sl4-2002-06-23-yudkowsky-0206-4208","name":"Re: Threats to the Singularity."},{"id":"ai-existential-risk-in-economic-analysis","name":"Existential risk from AI as a matter for economic analysis"},{"id":"gans-2018-self-regulating-agi","name":"Self-Regulating Artificial General Intelligence"},{"id":"ai21-labs","name":"AI21 Labs"},{"id":"alexandr-wang","name":"Alexandr Wang"},{"id":"david-luan","name":"David Luan"},{"id":"karen-simonyan","name":"Karén Simonyan"},{"id":"moonshot-ai","name":"Moonshot AI / 月之暗面"},{"id":"openai-superalignment-team","name":"Superalignment team (OpenAI)"},{"id":"sogou","name":"Sogou / 搜狗"},{"id":"wang-xiaochuan","name":"Wang Xiaochuan"},{"id":"yoav-shoham","name":"Yoav Shoham"},{"id":"zhou-xinyu","name":"Zhou Xinyu"},{"id":"zico-kolter","name":"Zico Kolter"},{"id":"kimi-k2-thinking-card","name":"Kimi-K2-Thinking model card"},{"id":"g7-hiroshima-code-of-conduct-2023","name":"Hiroshima Process International Code of Conduct for Organizations Developing Advanced AI Systems"},{"id":"us-bureau-of-industry-and-security","name":"Bureau of Industry and Security, U.S. Department of Commerce"},{"id":"us-bis-foundry-dd-ifr-2025","name":"Implementation of Additional Due Diligence Measures for Advanced Computing Integrated Circuits (interim final rule)"},{"id":"uk-aisi-evaluations-may-2024","name":"Advanced AI evaluations at AISI: May update"}],"claims":[{"id":"issued-fsf-3-1","statement":"Google DeepMind published Frontier Safety Framework v3.1, identified by its publication date of 2026-04-17.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Publication means a public document expression exists. The date is labelled publication; publication is not evidence that its requirements were implemented or effective in practice.","evidenceIds":["evidence-issued-fsf-3-1"]},{"id":"rsp-v2-2-insider-scope","statement":"RSP v2.2 excludes sophisticated insiders and state-compromised insiders from the ASL-3 Security Standard.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"This concerns the ASL-3 threat-model scope. It does not establish what protections Anthropic actually deployed or an overall safety change.","evidenceIds":["evidence-rsp-v2-2-insider-scope"],"attributes":{"changeBasis":"changelog"}},{"id":"fsf-v2-security-levels","statement":"FSF v2.0 introduces recommended security levels for Critical Capability Levels.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The source calls these recommendations and discusses collective adoption. This claim does not assert that every level was implemented.","evidenceIds":["evidence-fsf-v2-security-levels"],"attributes":{"changeBasis":"publisher_statement"}},{"id":"openai-abilene-use-2025","statement":"OpenAI reported early training and inference workloads at the partly operating Stargate I site in Abilene.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Partial operation does not establish that the whole site, or 4.5 GW expansion, was operational. Use is not ownership.","evidenceIds":["evidence-openai-abilene-use-2025"],"attributes":{"transactionType":"infrastructure_use","relationshipContext":"external","transactionStage":"operational"}},{"id":"google-anthropic-equity-2024","statement":"The CMA reported that Google held non-voting Anthropic shares as of August 2024.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Amount and percentage are redacted. The decision uses Google as shorthand for Alphabet’s group; a specific subscribing legal entity, voting control and later holdings are not inferred.","evidenceIds":["evidence-google-anthropic-equity-2024"],"attributes":{"transactionType":"equity_investment","relationshipContext":"external","transactionStage":"invested"}},{"id":"lineage-ii-leike-reward","statement":"Scalable agent alignment via reward modeling credits Jan Leike as a co-author.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Authorship is distinct from an employment interval and from institutional adoption of the proposed agenda.","evidenceIds":["evidence-lineage-ii-leike-reward"]},{"id":"atlas-openai-lp","statement":"OpenAI published its announcement creating a capped-profit partnership, OpenAI LP.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Publisher structural announcement, not independent verification of legal instruments. OpenAI is the umbrella atlas identity; the LP is not equated with the nonprofit.","evidenceIds":["evidence-atlas-openai-lp"]},{"id":"prov-rsp-v3-1-competitor-lead-delay","statement":"RSP v3.1 states that, in the scenario where Anthropic is in the lead, it will require a strong argument that catastrophic risk is contained, along the lines of its industry-wide recommendations, and will delay AI development and deployment as needed to achieve this until and unless it no longer believes it has a significant lead.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"What the edition states, quoted from the cited expression. It is not evidence of implementation, compliance or effectiveness.","evidenceIds":["evidence-prov-rsp-v3-1-competitor-lead-delay"],"attributes":{"provision":{"kind":"required_action","key":"rsp-competitor-lead-delay"}}},{"id":"prov-rsp-v2-2-may-proceed-with-safeguards","statement":"RSP v2.2 states that if, after its evaluations, Anthropic determines that it has met the ASL-3 Required Safeguards, it may proceed with deploying and training models above the Capability Threshold, provided it has also conducted a follow-up capability assessment.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"What the edition states, quoted from the cited expression. It is not evidence of implementation, compliance or effectiveness.","evidenceIds":["evidence-prov-rsp-v2-2-may-proceed-with-safeguards"],"attributes":{"provision":{"kind":"exception","key":"rsp-may-proceed-with-safeguards"}}},{"id":"prov-rsp-v2-0-policy-change-approval","statement":"RSP v2.0 states that changes to the policy will be proposed by the CEO and the Responsible Scaling Officer and approved by the Board of Directors, in consultation with the Long-Term Benefit Trust.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"What the edition states, quoted from the cited expression. It is not evidence of implementation, compliance or effectiveness.","evidenceIds":["evidence-prov-rsp-v2-0-policy-change-approval"],"attributes":{"provision":{"kind":"decision_authority","key":"rsp-policy-change-approval","parties":[{"role":"recommends","actor":"CEO"},{"role":"recommends","actor":"Responsible Scaling Officer"},{"role":"decides","actor":"Board of Directors"},{"role":"consulted","actor":"Long-Term Benefit Trust"}]}}},{"id":"absent-rsp-v2-0-weights-deletion-decision","statement":"No passage stating the provision “CISO and CEO convened to decide on weight deletion” was found in RSP · v2.0 within the inspected scope.","derivation":"inferred","resolution":"supported","inferenceRationale":"Premises: RSP · 2023 states this provision. The inspected scope of RSP · v2.0 was: RSP v2.0 (effective October 15, 2024), whole text layer, PDF pp. 1–22: cover, summary and contents (pp. 1–4); Introduction and §§1–7 (pp. 5–17); Appendices A–C (pp. 18–20); Changelog (pp. 21–22). Method: Read §3.3, §4.3, §6 and §7 in full and searched the text layer case-insensitively for CISO, CEO, convene, delet, weights, lock down, post-mortem, next ASL and Containment. No passage stating that the CISO and CEO, or other stakeholders, are to be convened to determine whether the level of danger merits deletion of the weights of a model that has surpassed a capability level during training was found. Hits examined and not counted: §6.2 'Stronger restrictions' (PDF p. 15) states that, if interim measures cannot adequately mitigate risk, in the security context Anthropic will delete model weights, without naming who decides; §6.2 'Interim measures' (PDF p. 14) lets the CEO and RSO approve interim measures; §3.3 and §4.3 (PDF pp. 10, 13) have the CEO and RSO make capability and safeguards determinations; §7.1 item 1 (PDF p. 15) lists RSO duties including approving training or deployment decisions. The conclusion is bounded by that scope and method.","editorialState":"seed","qualification":"A recorded absence within the stated scope: not a claim that the provision was deleted, and not a statement about practice or intent.","evidenceIds":["evidence-absent-rsp-v2-0-weights-deletion-decision"],"attributes":{"provision":{"kind":"decision_authority","key":"rsp-weights-deletion-decision"}}},{"id":"ts-b2-bostrom-1996-09-03","statement":"Bostrom’s message of 1996-09-03 asks whether a superhuman machine could not manipulate human society by subtle, reasonable-seeming suggestions while looking all the time like a perfectly decent machine.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A deployed AI in society, premised on there being only one such AI; no takeoff speed is asserted. \"wellfare\" is the message’s spelling.","evidenceIds":["evidence-ts-b2-bostrom-1996-09-03"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction"}}},{"id":"yk-y19-sl4-2002-06-23-yudkowsky-0206-4208-c","statement":"The SL4 message “Re: Threats to the Singularity.” (2002-06-23) contests Y19. Replying to Ben Goertzel, he separates deliberate extermination, which he thinks unlikely, from indifference, which he says would very probably have the same result.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: An AI would harm humans only out of hostility. Same passage as recorded for ‘A superintelligence would leave Earth alone’ (Y18). The argument that indifference equates to extermination is attributed to Eugen Leitl and endorsed; Leitl's message is not in the cache. Earlier than ‘Artificial Intelligence as a Positive and Negative Factor in Global Risk’ (2008). Quotation shortened to respect the ~10% cap for a short message.","evidenceIds":["evidence-yk-t19-s1"],"attributes":{"thought":{"roles":["contesting"],"register":"nonfiction","engagement":{"kind":"contrary-only","rationale":"Under the message-level rule, the passage states the contrary conclusion without giving a reason against the premise that an AI would harm humans only out of hostility, even when read with its message or dialogue turn."}}}},{"id":"yk-y27-agi-ruin-2022-c","statement":"The post “AGI Ruin: A List of Lethalities” (2022-06-05) contests Y27. He argues that alignment trained under safe conditions must generalise across a large distributional shift to dangerous conditions, because dangerous systems cannot be trained by trial and observed failure.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: Results from safe experiments on weaker systems will carry over to dangerous ones.","evidenceIds":["evidence-yk-t27-c"],"attributes":{"thought":{"roles":["contesting"],"register":"nonfiction","engagement":{"kind":"engages-premise","rationale":"Under the message-level rule, the passage, read with the paragraph, message or dialogue turn it belongs to, gives a reason against the premise that results from safe experiments on weaker systems will carry over to dangerous ones."}}}},{"id":"ec-e03-gans-2018-self-regulating-agi-a","statement":"Gans, “Self-Regulating Artificial General Intelligence” (2018-02) articulates E03. Gans takes Bostrom's control problem and \"paperclip apocalypse\" into economic theory. Gans states the scope of his self-regulation result.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"(First passage) Gans's own framing. (Second passage) A model result: it treats destruction of humanity as a possibility and argues that the model narrows it.","evidenceIds":["evidence-ec-e03-07","evidence-ec-e03-08"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","qualifier":"A model scenario under stated conditions, not the author’s own claim.","tier":{"value":1,"basis":"NBER working paper (core venue); Gans is an economist."},"conditions":"Assumptions A1 (an AI self-improves only by employing AIs with targeted goals) and A2 (improving power requires power-accumulation capabilities)."}}},{"id":"lg-gc-jingren-zhou-alibaba-joined-2015","statement":"Alibaba’s news site Alizila stated on 23 January 2025 that Jingren Zhou joined Alibaba Group in 2015.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Records the year of joining only; the article’s account of his earlier roles is not recorded.","evidenceIds":["lg-ev-gc-jingren-zhou-alibaba-joined-2015"],"attributes":{"role":"joined Alibaba Group (no title given)"}},{"id":"lg-deal-simonyan-inflection-chief-scientist-2024","statement":"Microsoft's 19 March 2024 announcement described Karén Simonyan as a co-founder and Chief Scientist of Inflection.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Microsoft's description at the moment Simonyan was announced as joining Microsoft; the record does not establish when the Inflection role ended.","evidenceIds":["lg-ev-deal-simonyan-inflection-chief-scientist-2024"],"attributes":{"role":"Co-founder and Chief Scientist"}},{"id":"lg-gs-superalignment-team-absorbed-2024","statement":"WIRED reported on 17 May 2024 that OpenAI had confirmed that its Superalignment team was no more; WIRED wrote that the group's work would be absorbed into OpenAI's other research efforts.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The record rests on reporting (WIRED), which relays OpenAI's confirmation that the team was no more; the sentence about the work being absorbed is WIRED's own, not attributed to OpenAI; no OpenAI announcement was retrieved. The record documents a team ended as a unit and its work absorbed into the rest of OpenAI; the record does not establish where each member went or how the work was reorganised.","evidenceIds":["lg-ev-gs-superalignment-team-absorbed-2024-a","lg-ev-gs-superalignment-team-absorbed-2024-b"]},{"id":"lg-oai-kolter-board-2024","statement":"OpenAI announced on 8 August 2024 the appointment of Zico Kolter to its board of directors.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The announcement does not name the legal entity; the record assigns the board to OpenAI, Inc., the nonprofit that OpenAI’s 2019 note names as OpenAI Inc and that, according to OpenAI’s structure page, has always governed and controlled the for-profit (context evidence).","evidenceIds":["lg-ev-oai-kolter-board-2024","lg-ev-oai-kolter-board-2024-inc","lg-ev-oai-kolter-board-2024-governed"],"attributes":{"role":"Board member"}},{"id":"lg-deal-luan-amazon-agi-sf-2024","statement":"Amazon Science identified David Luan on 9 December 2024 as VP of Autonomy and head of Amazon's AGI SF Lab.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A dated snapshot; it does not establish when Luan joined Amazon or whether he still holds the role.","evidenceIds":["lg-ev-deal-luan-amazon-agi-sf-2024"],"attributes":{"role":"VP of Autonomy and head of Amazon's AGI SF Lab"}},{"id":"lg-mx-wang-joins-meta-2025","statement":"Scale AI announced on 12 June 2025 that its founder Alexandr Wang was joining Meta to work on Meta's AI efforts.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Scale's announcement of a prospective move; it gives no Meta title. The title is recorded separately from Zuckerberg's 30 June memo.","evidenceIds":["lg-ev-mx-wang-joins-meta-2025"],"attributes":{"activityStage":"announced","role":"Joining Meta to work on Meta's AI efforts"}},{"id":"lg-mx-xai-into-spacex-announced-2026","statement":"Elon Musk wrote on 6 May 2026 that xAI would be dissolved as a separate company and continue as SpaceXAI, the AI products from SpaceX.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A stated plan in a reply post by Musk; it does not establish a legal dissolution or merger, or its date.","evidenceIds":["lg-ev-mx-xai-into-spacex-announced-2026"],"attributes":{"activityStage":"announced"}},{"id":"lg-cn-wang-xiaochuan-sogou-ceo-prior","statement":"Baichuan Intelligence’s website describes Wang Xiaochuan as former CEO of Sogou.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A description on Baichuan’s page, not a Sogou record; it establishes neither the tenure nor its end. English rendering of the Chinese passage is a machine translation; no human translation review.","evidenceIds":["lg-ev-cn-wang-xiaochuan-sogou-ceo-prior"],"attributes":{"role":"CEO (former, as described in 2026)"}},{"id":"lg-gc-zhou-xinyu-moonshot-cofounder","statement":"36Kr reported on 11 November 2025 that Zhou Xinyu is a co-founder of Moonshot AI.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The record rests on reporting; no Moonshot AI page naming its co-founders was retrieved. English rendering of the Chinese passage is a machine translation; no human translation review.","evidenceIds":["lg-ev-gc-zhou-xinyu-moonshot-cofounder"]},{"id":"lg-new-shoham-ai21-cofounder","statement":"AI21 Labs’ company page lists Yoav Shoham among its founders as co-founder.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The page lists two founders; Calcalist (reporting) names Amnon Shashua as a third. Calcalist’s report also names this person as a co-founder.","evidenceIds":["lg-ev-new-shoham-ai21-cofounder","lg-ev-new-shoham-ai21-cofounder-b"]},{"id":"ap-eliezer-yudkowsky-extropians-2000-09-05-yudkowsky-4292","statement":"The Extropians message “Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...)” of 2000-09-05, archive message 4292, is sent under the name Eliezer S. Yudkowsky.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A sender name on a mailing-list message as archived; it does not establish identity beyond the credited name, and quoted text by others inside the message is not the sender’s.","evidenceIds":["evidence-ap-eliezer-yudkowsky-extropians-2000-09-05-yudkowsky-4292"]},{"id":"cmp-ms-fgf-2025-d7-internal","statement":"Microsoft FGF · Feb 2025 states a provision on the comparison question “Internal reporting”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The edition also says employees can raise concerns through existing channels, anonymously if they choose, and are protected from retaliation (PDF p. 10). What the edition states, paraphrased from the cited expression with its page. It is not evidence of implementation, compliance or effectiveness, and a shared question does not make two editions’ provisions equivalent.","evidenceIds":["ev-cmp-ms-fgf-2025-d7-internal-1"],"attributes":{"provision":{"kind":"reporting","key":"comparison-d7-internal","facet":{"id":"D7.internal","status":"stated"}}}},{"id":"c2-moonshot-k2-thinking-card","statement":"Moonshot AI published the Kimi-K2-Thinking model on its moonshotai Hugging Face account; the model card, at a commit of 6 November 2025, describes it as a thinking model built starting from Kimi K2.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The card’s description is the developer’s own; benchmark claims are out of scope.","evidenceIds":["c2-ev-moonshot-k2-thinking-card"]},{"id":"fl-intl-g7-code-scope-voluntary","statement":"The Hiroshima Process Code of Conduct states that it will provide voluntary guidance for actions by organisations developing the most advanced AI systems, which it defines to include the most advanced foundation models and generative AI systems.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Defines the code's addressees by description only; it sets no compute or capability threshold.","evidenceIds":["evidence-fl-intl-g7-code-scope-voluntary-1"],"attributes":{"provision":{"kind":"threshold","key":"intl-g7-code-scope-voluntary"}}},{"id":"cpo-ch-bis-foundry-dd-2025-published","statement":"The Bureau of Industry and Security published the interim final rule \"Implementation of Additional Due Diligence Measures for Advanced Computing Integrated Circuits\" in the Federal Register of 16 January 2025, effective that day, with compliance required from 31 January 2025, to assist foundries and OSAT companies in complying with the advanced computing IC provisions.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Records publication and stated dates. The rule’s preamble names the AI diffusion rule as “a rule related to this IFR”; that reference is recorded separately.","evidenceIds":["evidence-cpo-ch-bis-foundry-dd-2025-published-1","evidence-cpo-ch-bis-foundry-dd-2025-published-2","evidence-cpo-ch-bis-foundry-dd-2025-published-3"]},{"id":"ab-ev-uk-aisi-may2024-chembio-result","statement":"The UK institute's May 2024 evaluations update states that several of the five publicly available large language models it tested demonstrated expert-level knowledge of chemistry and biology, answering over 600 private expert-written chemistry and biology questions at levels similar to humans with PhD-level training.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The result is the institute's own, for five models it does not name, its private question set and its stated human comparison. The post states that its work does not provide any assurance that a model is safe or unsafe, and that there may be a gap between evaluation performance and behaviour in the wild. Nothing is generalised to other models or to real-world capability. The expression checked is the copy served on 2026-10-28, which is later than first publication: it carries a note to readers that the institute changed its name to the AI Security Institute on 14 February 2025, and the address redirects to /blog/advanced-ai-evaluations-may-update. The record keeps the byline date of 20 May 2024, and the quoted wording was read in that served copy. The post's description of its question areas is summarised only at category level; no question or subject matter is recorded.","evidenceIds":["evidence-ab-ev-uk-aisi-may2024-chembio-result-1","evidence-ab-ev-uk-aisi-may2024-chembio-result-2","evidence-ab-ev-uk-aisi-may2024-chembio-result-3"],"attributes":{"evaluatedSystem":"Five publicly available large language models, anonymised in the post by colour","evaluation":{"what":"Whether the models could provide expert-level knowledge in chemistry and biology that could be used for beneficial but also harmful purposes","how":"Over 600 private expert-written open-text questions across four question areas, graded against the performance of human experts with relevant doctorates or work experience who had web access and up to an hour per question","result":"Several models demonstrated expert-level knowledge of chemistry and biology; most models performed similarly to the human experts, with two stated exceptions","caveats":"The post states that its work does not provide any assurance that a model is safe or unsafe, and that the institute remains aware of the potential gap between evaluation performance and performance in the wild."}}}],"sources":[{"id":"source-fsf-v2-update","title":"Updating the Frontier Safety Framework","authors":["Google DeepMind"],"publisher":"Google DeepMind","publicationDate":"2025-02-04","version":"Dated official update announcement","canonicalUrl":"https://deepmind.google/blog/updating-the-frontier-safety-framework/","evidenceUrl":"https://deepmind.google/blog/updating-the-frontier-safety-framework/","language":"en","archivedUrl":"https://web.archive.org/web/20251105050524/https://deepmind.google/blog/updating-the-frontier-safety-framework/","archivedAt":"2025-11-05","accessedAt":"2026-09-18","reuse":{"attribution":"Google DeepMind, Updating the Frontier Safety Framework.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-rsp-2-0","title":"Responsible Scaling Policy — v2.0","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":null,"version":"Responsible Scaling Policy v2.0; effective date 2024-10-15. v2.0 identity is supplied by the official archive and later dated changelogs; the cover is not numbered.","versionDate":"2024-10-15","canonicalUrl":"https://www-cdn.anthropic.com/616dee633636e5bd309cb73aed8622e80fe47839.pdf","evidenceUrl":"https://www-cdn.anthropic.com/616dee633636e5bd309cb73aed8622e80fe47839.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20241022154907/https://assets.anthropic.com/m/24a47b00f10301cd/original/Anthropic-Responsible-Scaling-Policy-2024-10-15.pdf","archivedAt":"2024-10-22","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v2.0.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-rsp-2-2","title":"Responsible Scaling Policy — v2.2","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":null,"version":"Responsible Scaling Policy v2.2; effective date 2025-05-14.","versionDate":"2025-05-14","canonicalUrl":"https://www-cdn.anthropic.com/872c653b2d0501d6ab44cf87f43e1dc4853e4d37.pdf","evidenceUrl":"https://www-cdn.anthropic.com/872c653b2d0501d6ab44cf87f43e1dc4853e4d37.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20250514223759/https://www-cdn.anthropic.com/872c653b2d0501d6ab44cf87f43e1dc4853e4d37.pdf","archivedAt":"2025-05-14","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v2.2.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-rsp-3-1","title":"Responsible Scaling Policy — v3.1","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":null,"version":"Responsible Scaling Policy v3.1; effective date 2026-04-02.","versionDate":"2026-04-02","canonicalUrl":"https://www-cdn.anthropic.com/files/4zrzovbb/website/bf04581e4f329735fd90634f6a1962c13c0bd351.pdf","evidenceUrl":"https://www-cdn.anthropic.com/files/4zrzovbb/website/bf04581e4f329735fd90634f6a1962c13c0bd351.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260407193305/https://www-cdn.anthropic.com/files/4zrzovbb/website/bf04581e4f329735fd90634f6a1962c13c0bd351.pdf","archivedAt":"2026-04-07","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v3.1.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-fsf-3-1","title":"Frontier Safety Framework — v3.1","authors":["Google DeepMind"],"publisher":"Google DeepMind","publicationDate":"2026-04-17","version":"Frontier Safety Framework v3.1; publication date 2026-04-17.","versionDate":"2026-04-17","canonicalUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf","evidenceUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260423021420/https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf","archivedAt":"2026-04-23","accessedAt":"2026-09-18","reuse":{"attribution":"Google DeepMind, Frontier Safety Framework — v3.1.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-oracle-stargate-2025","title":"Stargate advances with 4.5 GW partnership with Oracle","authors":["OpenAI"],"publisher":"OpenAI","publicationDate":"2025-07-22","version":"Dated announcement inspected 2026-09-19; a historical company statement, not independent verification of implementation. Archived copy: the Wayback Machine capture of 2025-07-22 11:51:19 UTC, the earliest same-day capture found by a bounded search, read by a cross-family machine check on 2026-10-26.","canonicalUrl":"https://openai.com/index/stargate-advances-with-partnership-with-oracle/","evidenceUrl":"https://openai.com/index/stargate-advances-with-partnership-with-oracle/","language":"en","archivedUrl":"https://web.archive.org/web/20250722115119/https://openai.com/index/stargate-advances-with-partnership-with-oracle/","archivedAt":"2025-07-22","accessedAt":"2026-09-19","reuse":{"attribution":"OpenAI, Stargate advances with 4.5 GW partnership with Oracle (2025-07-22).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and editorial paraphrases only; no full source redistributed."}},{"id":"source-cma-google-anthropic-2024","title":"Alphabet Inc.’s partnership with Anthropic PBC — Decision on relevant merger situation","authors":["Competition and Markets Authority"],"publisher":"Competition and Markets Authority","publicationDate":"2024-12-24","version":"ME/7108/24, decision issued 19 November 2024; public redacted full text published 24 December 2024. Amounts and equity percentages are redacted.","canonicalUrl":"https://www.gov.uk/cma-cases/alphabet-inc-google-llc-slash-anthropic-merger-inquiry","evidenceUrl":"https://assets.publishing.service.gov.uk/media/676959bae6ff7c8a1fde9d33/Full_text_decision__.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20250915141849/https://assets.publishing.service.gov.uk/media/676959bae6ff7c8a1fde9d33/Full_text_decision__.pdf","archivedAt":"2025-09-15","accessedAt":"2026-09-19","reuse":{"attribution":"Competition and Markets Authority, Alphabet Inc.’s partnership with Anthropic PBC — Decision on relevant merger situation (2024-12-24).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and editorial paraphrases only; no full source redistributed."}},{"id":"source-reward-modeling-paper-v1","title":"Scalable agent alignment via reward modeling: a research direction","authors":["Jan Leike","David Krueger","Tom Everitt","Miljan Martic","Vishal Maini","Shane Legg"],"publisher":"arXiv","publicationDate":"2018-11-19","version":"arXiv:1811.07871v1; first-submission date from version history. Later revisions are not substituted.","versionDate":"2018-11-19","canonicalUrl":"https://arxiv.org/abs/1811.07871","evidenceUrl":"https://arxiv.org/pdf/1811.07871v1","language":"en","accessedAt":"2026-09-25","reuse":{"attribution":"Jan Leike; David Krueger; Tom Everitt; Miljan Martic; Vishal Maini; Shane Legg, Scalable agent alignment via reward modeling: a research direction.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and brief attributed paraphrases only; source publications and personal contact details are not redistributed."}},{"id":"source-atlas-openai-lp","title":"OpenAI LP","authors":["Greg Brockman","Ilya Sutskever","OpenAI"],"publisher":"OpenAI","publicationDate":"2019-03-11","version":"Dated publisher page inspected 2026-09-28 and again 2026-10-04 at its current address; the quotes used from 2026-10-04 also match the Internet Archive capture of 2019-03-11 (16:15 UTC) at the original address, which uses straight quotation marks and an additional standfirst.","canonicalUrl":"https://openai.com/index/openai-lp/","evidenceUrl":"https://openai.com/index/openai-lp/","language":"en","archivedUrl":"https://web.archive.org/web/20190311161512/https://openai.com/blog/openai-lp/","archivedAt":"2019-03-11","accessedAt":"2026-10-04","reuse":{"attribution":"Greg Brockman; Ilya Sutskever; OpenAI. OpenAI LP.","licence":"No redistribution licence asserted.","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-ms-fgf-2025","title":"Frontier Governance Framework — Version 1 (February 2025)","authors":["Microsoft"],"publisher":"Microsoft","publicationDate":"2025-02","version":"Frontier Governance Framework, Version 1 (February 2025); its change log reads “8 February 2025 – First version”. The file served at the recorded address (last modified 2025-02-12) is byte-identical to the Internet Archive capture of 2025-02-26, the earliest located.","versionDate":"2025-02-08","canonicalUrl":"https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/final/en-us/microsoft-brand/documents/Microsoft-Frontier-Governance-Framework.pdf","evidenceUrl":"https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/final/en-us/microsoft-brand/documents/Microsoft-Frontier-Governance-Framework.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20250226161803/https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/final/en-us/microsoft-brand/documents/Microsoft-Frontier-Governance-Framework.pdf","archivedAt":"2025-02-26","accessedAt":"2026-10-03","reuse":{"attribution":"Microsoft, Frontier Governance Framework, Version 1 (February 2025).","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-openai-our-structure","title":"Our structure","authors":["OpenAI"],"publisher":"OpenAI","publicationDate":null,"version":"Undated web page describing the structure announced on October 28, 2025; inspected live 2026-10-04. The quoted sentences are identical in the Internet Archive capture of 2025-10-29 (17:22 UTC); the board list differs between the two.","canonicalUrl":"https://openai.com/our-structure/","evidenceUrl":"https://openai.com/our-structure/","language":"en","archivedUrl":"https://web.archive.org/web/20251029172207/https://openai.com/our-structure/","archivedAt":"2025-10-29","accessedAt":"2026-10-04","reuse":{"attribution":"OpenAI, Our structure (web page).","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-extropians-1996-09-03-bostrom","title":"Re: Thinking about the future... (Extropians mailing list)","authors":["Nick Bostrom"],"publisher":"Extropians mailing list","publicationDate":"1996-09-03","version":"Archive HTML as served, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search; quoted lines excluded when reading the author’s text. Archive at extropians.weidai.com.","canonicalUrl":"http://extropians.weidai.com/extropians.96/0721.html","evidenceUrl":"http://extropians.weidai.com/extropians.96/0721.html","language":"en","archivedUrl":"https://web.archive.org/web/20150427135909/http://extropians.weidai.com/extropians.96/0721.html","archivedAt":"2015-04-27","accessedAt":"2026-10-06","reuse":{"attribution":"Nick Bostrom, \"Re: Thinking about the future...\", Extropians mailing list, 1996-09-03.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-agi-ruin-2022","title":"AGI Ruin: A List of Lethalities","authors":["Eliezer Yudkowsky"],"publisher":"LessWrong and Alignment Forum","publicationDate":"2022-06-05","version":"Text checked in the post’s first stored version through the LessWrong GraphQL API on 2026-10-06.","canonicalUrl":"https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","evidenceUrl":"https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","language":"en","archivedUrl":"https://web.archive.org/web/20220605225854/https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","archivedAt":"2022-06-05","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky, \"AGI Ruin: A List of Lethalities\", LessWrong and Alignment Forum, 2022-06-05.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-extropians-2000-09-05-yudkowsky-4292","title":"Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...) (Extropians mailing list)","authors":["Eliezer Yudkowsky"],"publisher":"Extropians mailing list","publicationDate":"2000-09-05","version":"Archive HTML as served, retrieved 2026-10-07 (SHA-256 028f05c2f60379be3dca62fb8ed1158c95f4a7b64cc00e5087e863a1708ebb91); quoted lines excluded when reading the author’s text. The served archive shows later URL rewriting and does not authenticate historical bytes. Archive at extropians.weidai.com.","canonicalUrl":"http://extropians.weidai.com/extropians.3Q00/4292.html","evidenceUrl":"http://extropians.weidai.com/extropians.3Q00/4292.html","language":"en","archivedUrl":"https://web.archive.org/web/20160414013212/http://extropians.weidai.com/extropians.3Q00/4292.html","archivedAt":"2016-04-14","accessedAt":"2026-10-07","reuse":{"attribution":"Eliezer Yudkowsky, “Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...)”, Extropians mailing list, 2000-09-05.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-sl4-2002-06-23-yudkowsky-0206-4208","title":"Re: Threats to the Singularity. (SL4 mailing list)","authors":["Eliezer Yudkowsky"],"publisher":"SL4 mailing list","publicationDate":"2002-06-23","version":"Archive HTML as served, retrieved 2026-10-07 (SHA-256 b33257954ad3d0323662b61808df93cd39decfb44b6833d3ee2885d25c37f8de); quoted lines excluded when reading the author’s text. The served archive shows later URL rewriting and does not authenticate historical bytes. Archive at sl4.org.","canonicalUrl":"http://sl4.org/archive/0206/4208.html","evidenceUrl":"http://sl4.org/archive/0206/4208.html","language":"en","archivedUrl":"https://web.archive.org/web/20260114045509/http://sl4.org/archive/0206/4208.html","archivedAt":"2026-01-14","accessedAt":"2026-10-07","reuse":{"attribution":"Eliezer Yudkowsky, “Re: Threats to the Singularity.”, SL4 mailing list, 2002-06-23.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-gans-2018-self-regulating-agi","title":"Self-Regulating Artificial General Intelligence","authors":["Joshua S. Gans"],"publisher":"NBER Working Paper 24352","publicationDate":"2018-02","version":"Current copy (NBER PDF), no archive capture compared. Retrieved 2026-10-11. SHA-256 80c9e1edf896eda6d385f65d3a656ce31ebc61c8aed053eeff52baf4b8f4ec8a.","canonicalUrl":"https://www.nber.org/system/files/working_papers/w24352/w24352.pdf","evidenceUrl":"https://www.nber.org/system/files/working_papers/w24352/w24352.pdf","language":"en","accessedAt":"2026-10-11","reuse":{"attribution":"Joshua S. Gans, “Self-Regulating Artificial General Intelligence”, NBER Working Paper 24352, 2018-02.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-cn-baichuan-home","title":"百川大模型-百川智能 (homepage, 关于我们 section)","authors":["Baichuan Intelligence"],"publisher":"Baichuan Intelligence (百川智能)","publicationDate":null,"version":"Undated page, retrieved on 2026-10-15 (SHA-256 5740caddd43fdb073ceaf5f4737fae35086e93935f0e60ea45302901acd33b4a) through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.baichuan-ai.com/","evidenceUrl":"https://www.baichuan-ai.com/","language":"zh","archivedUrl":"https://web.archive.org/web/20261007164351/https://www.baichuan-ai.com/","archivedAt":"2026-10-07","accessedAt":"2026-10-15","reuse":{"attribution":"Baichuan Intelligence, “百川大模型-百川智能 (homepage, 关于我们 section)”.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-deal-amazon-agi-sf-2024","title":"Amazon opens new AI lab in San Francisco focused on long-term research bets","authors":["David Luan","Pieter Abbeel"],"publisher":"Amazon Science","publicationDate":"2024-12-09","version":"Dated page, retrieved on 2026-10-15 (SHA-256 112d55dc9d320457b9ab0e057ffcd0c2e2fa71e2fa44a40d345882f91204177a) directly from the publisher.","canonicalUrl":"https://www.amazon.science/blog/amazon-opens-new-ai-lab-in-san-francisco-focused-on-long-term-research-bets","evidenceUrl":"https://www.amazon.science/blog/amazon-opens-new-ai-lab-in-san-francisco-focused-on-long-term-research-bets","language":"en","archivedUrl":"https://web.archive.org/web/20260627193644/https://www.amazon.science/blog/amazon-opens-new-ai-lab-in-san-francisco-focused-on-long-term-research-bets","archivedAt":"2026-06-27","accessedAt":"2026-10-15","reuse":{"attribution":"David Luan, Pieter Abbeel, “Amazon opens new AI lab in San Francisco focused on long-term research bets”, 2024-12-09.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-deal-msft-suleyman-2024","title":"Mustafa Suleyman, DeepMind and Inflection Co-founder, joins Microsoft to lead Copilot","authors":["Satya Nadella"],"publisher":"Microsoft","publicationDate":"2024-03-19","version":"Dated page, retrieved on 2026-10-15 (SHA-256 45613ab2bd14157eef4a4ebe9eedcb47d84a36058b3ba099115c64844dc327ad) directly from the publisher.","canonicalUrl":"https://blogs.microsoft.com/blog/2024/03/19/mustafa-suleyman-deepmind-and-inflection-co-founder-joins-microsoft-to-lead-copilot/","evidenceUrl":"https://blogs.microsoft.com/blog/2024/03/19/mustafa-suleyman-deepmind-and-inflection-co-founder-joins-microsoft-to-lead-copilot/","language":"en","archivedUrl":"https://web.archive.org/web/20261014225026/https://blogs.microsoft.com/blog/2024/03/19/mustafa-suleyman-deepmind-and-inflection-co-founder-joins-microsoft-to-lead-copilot/","archivedAt":"2026-10-14","accessedAt":"2026-10-15","reuse":{"attribution":"Satya Nadella, “Mustafa Suleyman, DeepMind and Inflection Co-founder, joins Microsoft to lead Copilot”, 2024-03-19.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-gc-36kr-moonshot-2025","title":"K2 Thinking再炸场，杨植麟凌晨回答了21个问题","authors":["邓咏仪"],"publisher":"36氪 (36Kr)","publicationDate":"2025-11-11","version":"Dated page, retrieved on 2026-10-15 (SHA-256 a16573df2a18ca588067cb71a2ffb383e12a10c8461395e63e0165bf9aee12ae) directly from the publisher.","canonicalUrl":"https://www.36kr.com/p/3548523752173447","evidenceUrl":"https://www.36kr.com/p/3548523752173447","language":"zh","archivedUrl":"https://web.archive.org/web/20260415154308/https://www.36kr.com/p/3548523752173447","archivedAt":"2026-04-15","accessedAt":"2026-10-15","reuse":{"attribution":"邓咏仪, “K2 Thinking再炸场，杨植麟凌晨回答了21个问题”, 2025-11-11.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-gc-alizila-zhou-2025","title":"Jingren Zhou Recognized as ACM Fellow for Pioneering Contributions to Cloud Computing and Query Processing","authors":["Selina Zhang"],"publisher":"Alizila (Alibaba Group)","publicationDate":"2025-01-23","version":"Dated page, retrieved on 2026-10-15 (SHA-256 09d01e5964aff8b838ff8aa99eda59e6ad6deaecc6dd34206fdb9f5d4552e213) directly from the publisher.","canonicalUrl":"https://www.alizila.com/jingren-zhou-recognized-as-acm-fellow-for-pioneering-contributions-to-ai-and-cloud-computing/","evidenceUrl":"https://alizila.com/?p=64628","language":"en","archivedUrl":"https://web.archive.org/web/20251023205955/https://www.alizila.com/jingren-zhou-recognized-as-acm-fellow-for-pioneering-contributions-to-ai-and-cloud-computing/","archivedAt":"2025-10-23","accessedAt":"2026-10-15","reuse":{"attribution":"Selina Zhang, “Jingren Zhou Recognized as ACM Fellow for Pioneering Contributions to Cloud Computing and Query Processing”, 2025-01-23.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-gs-wired-superalignment-2024","title":"OpenAI’s Long-Term AI Risk Team Has Disbanded","authors":["Will Knight"],"publisher":"WIRED","publicationDate":"2024-05-17","version":"Dated page, retrieved on 2026-10-15 (SHA-256 0d45d5d10183fb40434e2cd12358744848d643be781a6758eb4046725341f6a2) directly from the publisher.","canonicalUrl":"https://www.wired.com/story/openai-superalignment-team-disbanded/","evidenceUrl":"https://www.wired.com/story/openai-superalignment-team-disbanded/","language":"en","archivedUrl":"https://web.archive.org/web/20261012225901/https://www.wired.com/story/openai-superalignment-team-disbanded/","archivedAt":"2026-10-12","accessedAt":"2026-10-15","reuse":{"attribution":"Will Knight, “OpenAI’s Long-Term AI Risk Team Has Disbanded”, 2024-05-17.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-mx-musk-xai-dissolve","title":"Reply of 6 May 2026 on xAI and SpaceXAI (X)","authors":["Elon Musk"],"publisher":"X","publicationDate":"2026-05-06","version":"Dated post, retrieved on 2026-10-15 (SHA-256 61fabd0099d3e8305894b1f6800e8e2b446c815a6589715246a37ab72f5c17d2) through X’s own syndication endpoint for the post’s text, which truncates long notes.","canonicalUrl":"https://x.com/elonmusk/status/2052105373621121284","evidenceUrl":"https://x.com/elonmusk/status/2052105373621121284","language":"en","archivedUrl":"https://web.archive.org/web/20260828020012/https://x.com/elonmusk/status/2052105373621121284","archivedAt":"2026-08-28","accessedAt":"2026-10-15","reuse":{"attribution":"Elon Musk, “Reply of 6 May 2026 on xAI and SpaceXAI (X)”, 2026-05-06.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-mx-scale-next-phase","title":"Scale AI Announces Next Phase of Company's Evolution","authors":["Scale AI"],"publisher":"Scale AI","publicationDate":"2025-06-12","version":"Dated page, retrieved on 2026-10-15 (SHA-256 750e681904b4d7b87a7d4f5253297f2e0a5296c7d4c91f75ff0a6668cf919db1) directly from the publisher.","canonicalUrl":"https://scale.com/blog/scale-ai-announces-next-phase-of-company-evolution","evidenceUrl":"https://scale.com/blog/scale-ai-announces-next-phase-of-company-evolution","language":"en","archivedUrl":"https://web.archive.org/web/20260924103944/https://scale.com/blog/scale-ai-announces-next-phase-of-company-evolution","archivedAt":"2026-09-24","accessedAt":"2026-10-15","reuse":{"attribution":"Scale AI, “Scale AI Announces Next Phase of Company's Evolution”, 2025-06-12.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-new-ai21-about","title":"About AI21 (company page, undated)","authors":["AI21 Labs"],"publisher":"AI21 Labs","publicationDate":null,"version":"Undated page, retrieved on 2026-10-15 (SHA-256 3a474e548a7ea862c779be07d317c7aabc4f733cd8118a354363b1bd8a0ce100) directly from the publisher. Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.ai21.com/about/","evidenceUrl":"https://www.ai21.com/about/","language":"en","archivedUrl":"https://web.archive.org/web/20261012071423/https://www.ai21.com/about/","archivedAt":"2026-10-12","accessedAt":"2026-10-15","reuse":{"attribution":"AI21 Labs, “About AI21 (company page, undated)”.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-new-calcalist-shashua","title":"Amnon Shashua to leave AI21 after nine years","authors":["Meir Orbach"],"publisher":"Calcalist (Ctech)","publicationDate":"2026-07-07","version":"Dated page, retrieved on 2026-10-15 (SHA-256 8f4c0b137e866cea36810276e520f3d9298b2b99eabc889f86bb6fccde1ab2ba) directly from the publisher.","canonicalUrl":"https://www.calcalistech.com/ctechnews/article/r1eabd5qfl","evidenceUrl":"https://www.calcalistech.com/ctechnews/article/r1eabd5qfl","language":"en","archivedUrl":"https://web.archive.org/web/20261014210052/https://www.calcalistech.com/ctechnews/article/r1eabd5qfl","archivedAt":"2026-10-14","accessedAt":"2026-10-15","reuse":{"attribution":"Meir Orbach, “Amnon Shashua to leave AI21 after nine years”, 2026-07-07.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-oai-kolter-2024","title":"Zico Kolter Joins OpenAI’s Board of Directors","authors":["OpenAI"],"publisher":"OpenAI","publicationDate":"2024-08-08","version":"Dated page, retrieved on 2026-10-15 (SHA-256 87fe6e28c87881f8cf6c27ca06226b68e0f386d73eb42d6c2844e2a159817d1a) through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy.","canonicalUrl":"https://openai.com/index/zico-kolter-joins-openais-board-of-directors/","evidenceUrl":"https://openai.com/index/zico-kolter-joins-openais-board-of-directors/","language":"en","archivedUrl":"https://web.archive.org/web/20260915160049/https://openai.com/index/zico-kolter-joins-openais-board-of-directors/","archivedAt":"2026-09-15","accessedAt":"2026-10-15","reuse":{"attribution":"OpenAI, “Zico Kolter Joins OpenAI’s Board of Directors”, 2024-08-08.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"c2-src-kimi-k2t-readme","title":"Kimi-K2-Thinking model card","authors":["Moonshot AI"],"publisher":"moonshotai (Hugging Face)","publicationDate":"2025-11-06","version":"Dated page, retrieved on 2026-10-27 (SHA-256 f5d3c6102da9a0939074aa403c0f17af95c92cd979412d10e69b92e1e695bf12) directly from the publisher. Version: 357b94aee9d50ec88e5e6dd9550fd7f957cb1baa. Model card at the Hugging Face commit \"Update LICENSE\" of 6 November 2025; the repository was initialised on 4 November 2025; later edits not inspected.","canonicalUrl":"https://huggingface.co/moonshotai/Kimi-K2-Thinking/raw/357b94aee9d50ec88e5e6dd9550fd7f957cb1baa/README.md","evidenceUrl":"https://huggingface.co/moonshotai/Kimi-K2-Thinking/raw/357b94aee9d50ec88e5e6dd9550fd7f957cb1baa/README.md","language":"en","archivedUrl":"https://web.archive.org/web/20261028223634/https://huggingface.co/moonshotai/Kimi-K2-Thinking/raw/357b94aee9d50ec88e5e6dd9550fd7f957cb1baa/README.md","archivedAt":"2026-10-28","accessedAt":"2026-10-27","reuse":{"attribution":"Moonshot AI, “Kimi-K2-Thinking model card”, 2025-11-06.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-g7-hiroshima-code-of-conduct-2023","title":"Hiroshima Process International Code of Conduct for Organizations Developing Advanced AI Systems","authors":["G7"],"publisher":"Ministry of Internal Affairs and Communications, Japan (Hiroshima AI Process site)","publicationDate":null,"version":"Retrieved on 2026-10-27 (SHA-256 14cfe36633502daa48231b49ca7f186c53a72add181b328047665763ecb13ac4). Undated English PDF; dated by the G7 Leaders' Statement of 30 October 2023, which attaches it","canonicalUrl":"https://www.soumu.go.jp/hiroshimaaiprocess/pdf/document05_en.pdf","evidenceUrl":"https://www.soumu.go.jp/hiroshimaaiprocess/pdf/document05_en.pdf","language":"en","accessedAt":"2026-10-27","reuse":{"attribution":"G7, “Hiroshima Process International Code of Conduct for Organizations Developing Advanced AI Systems”.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-cpo-ch-bis-foundry-dd-2025","title":"Implementation of Additional Due Diligence Measures for Advanced Computing Integrated Circuits; Amendments and Clarifications; and Extension of Comment Period (90 FR 5298)","authors":["Bureau of Industry and Security, U.S. Department of Commerce"],"publisher":"Bureau of Industry and Security, in the Federal Register (govinfo.gov)","publicationDate":"2025-01-16","version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 e1f7ddd6e4fd08d1ee1470bd001d38ca16adb28bb42444640775a63c7bef1026). Federal Register print of Thursday, January 16, 2025, Vol. 90, No. 10, beginning at p. 5298, FR Doc. 2025-00711 (filed 1-15-25). Interim final rule. Retrieved again at admission on 2026-10-30 (SHA-256 e1f7ddd6e4fd08d1ee1470bd001d38ca16adb28bb42444640775a63c7bef1026, the same bytes).","canonicalUrl":"https://www.govinfo.gov/content/pkg/FR-2025-01-16/pdf/2025-00711.pdf","evidenceUrl":"https://www.govinfo.gov/content/pkg/FR-2025-01-16/pdf/2025-00711.pdf","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"Bureau of Industry and Security, U.S. Department of Commerce, “Implementation of Additional Due Diligence Measures for Advanced Computing Integrated Circuits; Amendments and Clarifications; and Extension of Comment Period (90 FR 5298)”, 2025-01-16.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-ab-aisi-may2024","title":"Advanced AI evaluations at AISI: May update","authors":["AI Security Institute (UK), published under its earlier name AI Safety Institute"],"publisher":"AI Security Institute (UK), published under its earlier name AI Safety Institute","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 194d183711498e8fa0fa995229646e73b58c6cb7df533806585f373fbcd5f917). Blog post bylined 'Technical staff — May 20, 2024', as served on 2026-10-28 after a redirect to https://www.aisi.gov.uk/blog/advanced-ai-evaluations-may-update; it carries a later note that the institute changed its name on 14 February 2025. This is a different document from the dataset's uk-aisi-evaluations-2024 (the approach-to-evaluations publication of 9 February 2024). Retrieved again at admission on 2026-10-30 (SHA-256 194d183711498e8fa0fa995229646e73b58c6cb7df533806585f373fbcd5f917, the same bytes).","canonicalUrl":"https://www.aisi.gov.uk/work/advanced-ai-evaluations-may-update","evidenceUrl":"https://www.aisi.gov.uk/work/advanced-ai-evaluations-may-update","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"AI Security Institute (UK), published under its earlier name AI Safety Institute, “Advanced AI evaluations at AISI: May update”.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}}],"evidence":[{"id":"evidence-issued-fsf-3-1","sourceRole":"primary","support":"supports","locator":"PDF p. 1, cover: version and publication-date lines","passage":"Version 3.1 Published: April 17, 2026","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2026-04-23; it replaces an earlier attributed paraphrase. Human editorial review is pending."},{"id":"evidence-rsp-v2-2-insider-scope","sourceRole":"primary","support":"supports","locator":"PDF p. 23 (printed p. 19), Changelog, May 14, 2025 (RSP v2.2) entry, paragraph \"ASL-3 Security\", first two sentences","passage":"This update excludes both sophisticated insiders and state-compromised insiders from the ASL-3 Security Standard. Previously, only “highly sophisticated state-compromised insiders” were explicitly excluded.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2025-05-14; it replaces an earlier attributed paraphrase. Human editorial review is pending."},{"id":"evidence-fsf-v2-security-levels","sourceRole":"primary","support":"supports","locator":"Section \"Recommendations for Heightened Security\", second paragraph, first sentence","passage":"Since then, we have drawn on wider research to evolve these security mitigation levels and recommend a level for each of our CCLs.*","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the live page and the Internet Archive capture of 2025-11-05; it replaces an earlier attributed paraphrase. Human editorial review is pending."},{"id":"evidence-openai-abilene-use-2025","sourceRole":"primary","support":"supports","locator":"Paragraph beginning “Meanwhile, construction of Stargate I”","passage":"parts of the facility are now up and running. … we recently began running early training and inference workloads","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Re-checked 2026-10-15 (research-debt pass resolving Audit B's RESEARCH MORE). openai.com refused direct retrieval (HTTP 403) and the Wayback Machine did not answer; the quotation was verified verbatim against the current page text obtained through a reader relay (r.jina.ai), not against a dated capture, so it does not authenticate the July 2025 bytes. Company statements are not independent proof of implementation. Human source review pending."},{"id":"evidence-google-anthropic-equity-2024","sourceRole":"primary","support":"supports","locator":"Printed page 5, paragraph 16(a); page 1, paragraph 1 for group naming","passage":"The CMA records acquired non-voting shares and an August 2024 holding.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-19","note":"Exact cited source passage inspected by a machine agent. Company statements are not independent proof of implementation. Human source review pending."},{"id":"evidence-lineage-ii-leike-reward","sourceRole":"primary","support":"supports","locator":"PDF p. 1, author block","passage":"The author block includes Jan Leike with a DeepMind affiliation.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-25","note":"Direct source inspection; machine check only. Authorship is distinct from an employment interval and from institutional adoption of the proposed agenda."},{"id":"evidence-atlas-openai-lp","sourceRole":"primary","support":"supports","locator":"Introduction, paragraphs beginning “We want to increase” and “The fundamental idea”","passage":"The announcement introduces OpenAI LP, allowing capped investor and employee returns.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-28","note":"Original source passage checked by the machine agent that proposed the record. No independent machine audit or human source inspection is recorded for this increment."},{"id":"evidence-prov-rsp-v3-1-competitor-lead-delay","sourceRole":"primary","support":"supports","locator":"PDF p. 17 (printed p. 16), Appendix A: Commitments Related to Competitors, table row 'Anthropic in the lead', Commitment column","passage":"We will require a strong argument that catastrophic risk is contained, along the lines of our recommendations for industry-wide safety (see Section 1). We will delay AI development and deployment as needed to achieve this, until and unless we no longer believe we have a significant lead.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2026-04-07. Extraction note: Unchanged from v3.0. v3.1 adds two sentences to the Appendix A introduction (PDF p. 16) on cautionary action and considering pausing outside these scenarios; see the absence record. Human editorial review is pending."},{"id":"evidence-prov-rsp-v2-2-may-proceed-with-safeguards","sourceRole":"primary","support":"supports","locator":"PDF p. 14 (printed p. 10), §4.3 Safeguards Decision, first paragraph, first sentence","passage":"If, after the evaluations above, we determine that we have met the ASL-3 Required Safeguards, then we may proceed with deploying and training models above the Capability Threshold, provided we have also conducted a follow-up capability assessment.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2025-05-14. Extraction note: Wording identical to v2.1; no change proposed. In the v2.2 text layer the following sentence runs on without a space ('assessment.The process'); the quote ends before it. Human editorial review is pending."},{"id":"evidence-prov-rsp-v2-0-policy-change-approval","sourceRole":"primary","support":"supports","locator":"PDF p. 16 (printed p. 12), §7.1 Internal Governance, numbered item 7 ('Policy changes'), first sentence","passage":"Changes to this policy will be proposed by the CEO and the Responsible Scaling Officer and approved by the Board of Directors, in consultation with the Long-Term Benefit Trust.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2024-10-22 (earlier assets.anthropic.com address). Extraction note: Footnote marker 18 follows the sentence in the text layer ('Trust.18'); footnote 18 is the lower-safeguards-if-others-pass clause recorded separately. 'Proposed by' is mapped to the 'recommends' role. Human editorial review is pending."},{"id":"evidence-absent-rsp-v2-0-weights-deletion-decision","sourceRole":"primary","support":"supports","locator":"RSP v2.0 (effective October 15, 2024), whole text layer, PDF pp. 1–22: cover, summary and contents (pp. 1–4); Introduction and §§1–7 (pp. 5–17); Appendices A–C (pp. 18–20); Changelog (pp. 21–22).","passage":"Within the inspected scope, no passage stating the provision “CISO and CEO convened to decide on weight deletion” was found.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-03","note":"Absence recorded by an automated research assistant on 2026-10-03. Method: Read §3.3, §4.3, §6 and §7 in full and searched the text layer case-insensitively for CISO, CEO, convene, delet, weights, lock down, post-mortem, next ASL and Containment. No passage stating that the CISO and CEO, or other stakeholders, are to be convened to determine whether the level of danger merits deletion of the weights of a model that has surpassed a capability level during training was found. Hits examined and not counted: §6.2 'Stronger restrictions' (PDF p. 15) states that, if interim measures cannot adequately mitigate risk, in the security context Anthropic will delete model weights, without naming who decides; §6.2 'Interim measures' (PDF p. 14) lets the CEO and RSO approve interim measures; §3.3 and §4.3 (PDF pp. 10, 13) have the CEO and RSO make capability and safeguards determinations; §7.1 item 1 (PDF p. 15) lists RSO duties including approving training or deployment decisions. Human editorial review is pending."},{"id":"evidence-ts-b2-bostrom-1996-09-03","sourceRole":"primary","support":"supports","locator":"Message body, third paragraph (\"wellfare\" sic)","passage":"Could such a machine [...] not manipulate a human society by subtle suggestions that seem very reasonable [...]? And all the time it would look as if it were a perfectly decent machine","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Checked against the archive HTML on 2026-10-06 (hash matched the 1996–2022 candidate search)."},{"id":"evidence-yk-t19-s1","sourceRole":"primary","support":"supports","locator":"message body, paragraph beginning 'Ben, it is extremely unlikely that Powers'","passage":"extremely unlikely … exterminate us deliberately … indifference very probably equates to extermination.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-07","note":"Candidate pack passage T19-s1. The quotation (13 words; trimmed 2026-10-13 from 17 words) matched the retrieved expression (SHA-256 b33257954ad3d0323662b61808df93cd39decfb44b6833d3ee2885d25c37f8de) when checked on 2026-10-07."},{"id":"evidence-yk-t27-c","sourceRole":"primary","support":"supports","locator":"Section B.1 The distributional leap, item 10 (opening sentence)","passage":"You can't train alignment by running lethally dangerous cognitions, observing whether the outputs kill or deceive or corrupt the operators","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Candidate pack passage T27-c. The quotation (20 words; trimmed 2026-10-13 from 27 words) matched the retrieved expression (SHA-256 b36f38a6fd1020bfbd7b9d941e150ea68561242f3a41a0b8f2a56b853a9a6db3) when checked on 2026-10-06. It also matched the posting-day version 1.0.0. Pack version note: edited post, current version 1.3.0; posting-day check pending."},{"id":"evidence-ec-e03-07","sourceRole":"primary","support":"supports","locator":"NBER WP 24352 (February 2018), sec. 1 \"Introduction\", p. 2","passage":"This paper examines the control problem using the tools of economic theory. As has been noted by many, the control problem is a superintelligence version of the principal-agent problem","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-16","note":"Candidate pack passage E03-07 (research/economics-mainstream/esoteric). The quotation (29 words) matched the retrieved current copy (SHA-256 80c9e1edf896eda6d385f65d3a656ce31ebc61c8aed053eeff52baf4b8f4ec8a) in the pack's verification of 2026-10-15. Re-checked at admission on 2026-10-16 against a copy retrieved that day."},{"id":"evidence-ec-e03-08","sourceRole":"primary","support":"supports","locator":"NBER WP 24352, sec. 1, p. 3","passage":"To be sure, while this approach cannot rule out destruction of humanity by a superintelligent AGI, it dramatically limits the scope by which that might arise.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-16","note":"Candidate pack passage E03-08 (research/economics-mainstream/esoteric). The quotation (26 words) matched the retrieved current copy (SHA-256 80c9e1edf896eda6d385f65d3a656ce31ebc61c8aed053eeff52baf4b8f4ec8a) in the pack's verification of 2026-10-15. Re-checked at admission on 2026-10-16 against a copy retrieved that day.","contextOnly":{"reason":"A model scenario, not the author's own forecast (maintainer decision D3).","requires":"The model's conditions and the author's own assessment."}},{"id":"lg-ev-gc-jingren-zhou-alibaba-joined-2015","sourceRole":"primary","support":"supports","locator":"Paragraph beginning “Before joining Alibaba Cloud”","passage":"In 2015, Jingren Zhou joined Alibaba Group","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (7 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-deal-simonyan-inflection-chief-scientist-2024","sourceRole":"primary","support":"supports","locator":"Paragraph 5","passage":"Karén, a Co-founder and Chief Scientist of Inflection","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (8 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-gs-superalignment-team-absorbed-2024-a","sourceRole":"secondary","support":"supports","locator":"Paragraph 2","passage":"Now OpenAI’s “superalignment team” is no more, the company confirms.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (10 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-16/17) re-read the record’s sources and judged it partly supported; its findings were applied at admission. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-gs-superalignment-team-absorbed-2024-b","sourceRole":"secondary","support":"supports","locator":"Paragraph 2","passage":"The group’s work will be absorbed into OpenAI’s other research efforts.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (11 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-16/17) re-read the record’s sources and judged it partly supported; its findings were applied at admission. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-oai-kolter-board-2024","sourceRole":"primary","support":"supports","locator":"Opening sentence","passage":"the appointment of Zico Kolter to OpenAI’s Board of Directors","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (10 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-oai-kolter-board-2024-inc","sourceRole":"primary","support":"context","locator":"Paragraph defining the general partner","passage":"OpenAI’s 2019 note gives the legal name of OpenAI Nonprofit as OpenAI Inc.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “whose legal name is “OpenAI Inc”” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-oai-kolter-board-2024-governed","sourceRole":"primary","support":"context","locator":"Opening section (page retrieved on 15 October 2026)","passage":"The structure page states that the for-profit created in 2019 has always been governed and controlled by the nonprofit, now the OpenAI Foundation.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “governed and controlled by the nonprofit” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-deal-luan-amazon-agi-sf-2024","sourceRole":"primary","support":"supports","locator":"Photo caption","passage":"David Luan, VP of Autonomy and head of Amazon's AGI SF Lab","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (12 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-mx-wang-joins-meta-2025","sourceRole":"primary","support":"supports","locator":"Paragraph 2","passage":"Scale’s founder, Alexandr Wang, is joining Meta to work on Meta’s AI efforts.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-29","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (13 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review. On 2026-10-30, by the maintainer’s decision of that day, the quotation was corrected to the archive capture of 2026-09-24 (quotation marks and apostrophes as in the capture), which a cross-family archive check read on 2026-10-29; the proposition is unchanged. The retrieval route is that of the earlier check.","retrievalRoute":"direct"},{"id":"lg-ev-mx-xai-into-spacex-announced-2026","sourceRole":"primary","support":"supports","locator":"Reply post","passage":"Musk writes that xAI will be dissolved as a separate company and become simply SpaceXAI, the AI products from SpaceX.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “be dissolved” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through X’s own syndication endpoint for the post’s text, which truncates long notes. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"x_syndication"},{"id":"lg-ev-cn-wang-xiaochuan-sogou-ceo-prior","sourceRole":"primary","support":"supports","locator":"Homepage, 关于我们 (About us) section","passage":"百川智能成立于2023年3月24日，由前搜狗公司CEO王小川创立","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (1 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-16/17) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-gc-zhou-xinyu-moonshot-cofounder","sourceRole":"secondary","support":"supports","locator":"Section on the Reddit AMA, caption and quotation attribution","passage":"The report attributes a remark to “Moonshot AI co-founder Zhou Xinyu”.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “月之暗面联合创始人周昕宇表示” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. Paraphrase of a passage not in English; no human translation review. A cross-family machine check (2026-10-16/17) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-new-shoham-ai21-cofounder","sourceRole":"primary","support":"supports","locator":"Section “Our founders”","passage":"The founders section lists Yoav Shoham as chairman and co-founder.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “Prof. Yoav Shoham Chairman & Co-Founder” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-new-shoham-ai21-cofounder-b","sourceRole":"secondary","support":"supports","locator":"Paragraph 3","passage":"AI21 Labs was founded in 2017 by Prof. Yoav Shoham, Ori Goshen, and Prof. Amnon Shashua.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (16 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ap-eliezer-yudkowsky-extropians-2000-09-05-yudkowsky-4292","sourceRole":"primary","support":"supports","locator":"Message header, sender name (address not reproduced)","passage":"Eliezer S. Yudkowsky","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Credited name checked by an automated research assistant on 2026-10-27 in a direct retrieval. Human editorial review is pending.","retrievalRoute":"direct"},{"id":"ev-cmp-ms-fgf-2025-d7-internal-1","sourceRole":"primary","support":"supports","locator":"PDF p. 10 (printed p. 9), §5 Governance, Risk-informed deployment decision, first paragraph","passage":"The edition says documentation of the capability assessments before and after mitigation will go to the “Executive Officers” responsible for Microsoft's AI governance programme, or their delegates.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-27","note":"Explicit paraphrase, written on 2026-10-27 by an automated research assistant after re-reading the passage in the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. The comparison quoted the passage; it is given here as a paraphrase to keep the edition within the quotation limit. Human editorial review is pending."},{"id":"c2-ev-moonshot-k2-thinking-card","sourceRole":"primary","support":"supports","locator":"Model card, “1. Model Introduction”","passage":"Starting with Kimi K2, we built it as a thinking agent that reasons step-by-step while dynamically invoking tools.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Candidate pack passage, Chinese frontier-AI institutions, second collection (research/china-labs-2). The quotation (18 words) matched the retrieved text on 2026-10-27 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-intl-g7-code-scope-voluntary-1","sourceRole":"primary","support":"supports","locator":"PDF p. 1, first paragraph","passage":"will provide voluntary guidance for actions by organizations developing the most advanced AI systems, including the most advanced foundation models and generative AI systems (henceforth \"advanced AI systems\").","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (28 words) matched the text retrieved for the candidate pack on 2026-10-27 and again at admission on 2026-10-28, in a fresh retrieval whose SHA-256 equals the pack’s. Read directly from www.soumu.go.jp. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-ch-bis-foundry-dd-2025-published-1","sourceRole":"primary","support":"supports","locator":"90 FR 5298, masthead","passage":"Federal Register / Vol. 90, No. 10 / Thursday, January 16, 2025 / Rules and Regulations","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (13 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-29/30, item 70) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-ch-bis-foundry-dd-2025-published-2","sourceRole":"primary","support":"supports","locator":"90 FR 5298, DATES","passage":"The effective date of this rule is January 16, 2025.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (10 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-29/30, item 70) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-ch-bis-foundry-dd-2025-published-3","sourceRole":"primary","support":"supports","locator":"90 FR 5298, SUMMARY","passage":"assist foundries and Outsourced Semiconductor Assembly and Test (‘‘OSATs’’) companies in complying with provisions of the EAR pertaining to advanced computing ICs","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (22 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-29/30, item 70) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-ev-uk-aisi-may2024-chembio-result-1","sourceRole":"primary","support":"supports","locator":"'Here, we highlight a selection of recent results' list, first item","passage":"Several LLMs demonstrated expert-level knowledge of chemistry and biology. Models answered over 600 private expert-written chemistry and biology questions at similar levels to humans with PhD-level training.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (27 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-30, item 16) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-ev-uk-aisi-may2024-chembio-result-2","sourceRole":"primary","support":"context","locator":"'Outlook' section, final paragraph","passage":"Our work does not provide any assurance that a model is \"safe\" or \"unsafe\".","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (14 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. The institute's own limit on how its results may be read. Locator corrected after the audit of 2026-10-28: the post's closing section is headed 'Outlook', not 'limitations'. A cross-family machine check (2026-10-30, item 16) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-ev-uk-aisi-may2024-chembio-result-3","sourceRole":"primary","support":"context","locator":"'Outlook' section, penultimate paragraph","passage":"we remain acutely aware of the potential gap between how advanced AI systems perform in our evaluations versus how they may perform in the wild","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (25 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. Locator corrected after the audit of 2026-10-28 for the same reason; this paragraph precedes the one above. A cross-family machine check (2026-10-30, item 16) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"}],"at":{"entities":[0,1,2,3,5,19,22,39,42,43,44,46,50,52,62,74,101,121,149,157,179,182,201,207,235,289,333,343,368,375,390,469,587,597,616,620,657,717,747,759,799,820,834,842,845,965,1026,1064,1087,1251],"claims":[68,89,95,114,127,229,278,418,534,536,583,865,1117,1164,1523,1573,1680,1688,1718,1740,1765,1866,1947,1964,2030,2382,2715,2817,2930,3067,3367],"sources":[31,36,38,40,47,58,64,100,121,183,212,290,338,357,436,546,578,589,600,606,608,669,705,708,720,721,747,795,859,911,1089],"evidence":[71,106,112,133,146,250,299,450,600,602,650,1079,1357,1413,1898,1899,1970,2120,2132,2133,2182,2183,2184,2228,2271,2407,2512,2537,2624,2625,3029,3410,3592,3788,4020,4021,4022,4622,4623,4624]},"version":"1lg47t7"}