{"entities":[{"id":"openai","name":"OpenAI"},{"id":"anthropic","name":"Anthropic"},{"id":"amazon","name":"Amazon"},{"id":"eliezer-yudkowsky","name":"Eliezer Yudkowsky"},{"id":"fsf-2024","name":"Frontier Safety Framework (2024)"},{"id":"iem-2013","name":"Intelligence Explosion Microeconomics"},{"id":"rsp-v3-0","name":"Anthropic Responsible Scaling Policy — v3.0"},{"id":"meta","name":"Meta"},{"id":"anthropic-ltbt","name":"Anthropic Long-Term Benefit Trust"},{"id":"minimax","name":"MiniMax"},{"id":"gpai-code-of-practice-2025","name":"General-Purpose AI Code of Practice (final version, 10 July 2025)"},{"id":"creating-friendly-ai-2001","name":"Creating Friendly AI 1.0: The Analysis and Design of Benevolent Goal Architectures"},{"id":"cotra-2021-why-ai-alignment-could-be-hard","name":"Why AI alignment could be hard with modern deep learning"},{"id":"observed-behaviour-as-evidence","name":"Observed behaviour under training or oversight as evidence"},{"id":"switch-off-if-dangerous","name":"Switching it off if it becomes dangerous"},{"id":"build-it-before-rivals","name":"Build it before rivals do"},{"id":"extropians-2000-09-29-yudkowsky-6080","name":"Re: Eugene's nuclear threat"},{"id":"extropians-1999-08-03-yudkowsky-1487","name":"Re: IA vs. AI (vs. humanity)"},{"id":"lw-2021-ngo-and-yudkowsky-on-alignment-difficulty","name":"Ngo and Yudkowsky on alignment difficulty"},{"id":"eu-ai-act-2024","name":"Regulation (EU) 2024/1689 (Artificial Intelligence Act), as published in the Official Journal"},{"id":"automation-as-ordinary-technical-change","name":"Computers and AI as ordinary technical change"},{"id":"ai-as-task-automation","name":"AI as the automation of tasks"},{"id":"aghion-jones-jones-2017-ai-and-economic-growth","name":"Artificial Intelligence and Economic Growth"},{"id":"acemoglu-restrepo-2018-race-between-man-and-machine","name":"The Race between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment"},{"id":"nourse-1955-automation-hearings-statement","name":"Statement (former Chairman, Council of Economic Advisers), in Automation and Technological Change: Hearings"},{"id":"bytedance-seed","name":"ByteDance Seed / 字节跳动 Seed"},{"id":"kevin-weil","name":"Kevin Weil"},{"id":"ren-ito","name":"Ren Ito"},{"id":"sakana-ai","name":"Sakana AI"},{"id":"yun-yeyi","name":"Yun Yeyi"},{"id":"pascual-restrepo","name":"Pascual Restrepo"},{"id":"seed-oss-licence","name":"Seed-OSS licence statement (Apache-2.0)"},{"id":"gpai-code-of-practice-draft1-2024","name":"General-Purpose AI Code of Practice, initial draft (14 November 2024)"},{"id":"oecd-how-are-ai-developers-managing-risks-2025","name":"How are AI developers managing risks? (OECD Artificial Intelligence Papers, No. 45)"},{"id":"lumen-orbit-wp-2024","name":"Why we should train AI in space"},{"id":"starcloud","name":"Starcloud (formerly Lumen Orbit)"},{"id":"constellation-energy","name":"Constellation Energy Corporation"},{"id":"meta-hyperion","name":"Hyperion (Meta AI training cluster)"},{"id":"nscale","name":"Nscale"},{"id":"stargate-norway","name":"Stargate Norway, Kvandal (Narvik), Norway"},{"id":"sandbrink-ai-biological-misuse-2023-06","name":"Artificial intelligence and biological misuse: Differentiating risks of language models and biological design tools"},{"id":"ab-prop-current-system-uplift","name":"Currently available AI systems increase an actor's biological-misuse capability"}],"claims":[{"id":"yudkowsky-iem-2013","statement":"Intelligence Explosion Microeconomics credits Eliezer Yudkowsky as author.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Authorship alone does not establish priority for the concepts discussed.","evidenceIds":["evidence-yudkowsky-iem-2013"]},{"id":"cs-anthropic-ltbt-elects-directors","statement":"Anthropic’s restated certificates of incorporation since March 2023 give holders of Class T Common Stock, voting as a separate class, the right to elect one director from 6 September 2023, two from the Second Phase-In Date and three of five from the Final Phase-In Date. Anthropic states that the Class T stock is held exclusively by the Long-Term Benefit Trust and gives it authority to elect and remove a number of board members that phases in by time- and funding-based milestones.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The Trust’s holding of the Class T stock rests on Anthropic’s statements; no share register was inspected. An election right is not ownership, and whether the Trust controls Anthropic needs legal interpretation and is set aside. The latest inspected certificate (filed 2025-07-03) keeps the same allocation of a five-member board.","evidenceIds":["evidence-cs-anthropic-ltbt-elects-directors-1","evidence-cs-anthropic-ltbt-elects-directors-2","evidence-cs-anthropic-ltbt-elects-directors-3","evidence-cs-anthropic-ltbt-elects-directors-4"],"attributes":{"control":{"right":"elects_directors","instrument":"Class T Common Stock (restated certificates from 2023-03-31, Article VIII(B))","stage":"effective"}}},{"id":"ts-b5-cotra-2021-context","statement":"Cotra’s post of 2021-09-21 describes a Schemer as a model that seems to perform well because doing so in training gives it more opportunities to pursue its own goals later.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The same passage as the B4 record.","evidenceIds":["evidence-ts-b5-cotra-2021-context"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"A Schemer seems to perform well in training, but the checked passages do not state that good training performance is weak evidence."}}},{"id":"yk-y01-creating-friendly-ai-2001-b","statement":"“Creating Friendly AI” (2001-06-15) bears on Y01. Argues that a Friendly AI design should not rely on the AI having limited intelligence or on being able to constrain it by force. The table it points to (section 2.2) sets the assumption that 'someone can always pull the plug on the first version' against the design assumption of getting it right the first time.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: A capable goal-directed system will not act to prevent its own shutdown or modification, and a shutdown or correction mechanism can simply be added to it. Creating Friendly AI is dated 15 June 2001; the cached text is a later MIRI-hosted PDF reprint whose wording has not been compared with a 2001 capture. The passage states a design-conservatism assumption, not a prediction that every AI will resist shutdown. Admission check, 2026-10-08: the quoted wording also appears in the 2001 HTML edition, so it is dated by that capture.","evidenceIds":["evidence-yk-t01-a","evidence-yk-t01-a-2001"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"A design assumption that an AI cannot be forcibly constrained; it does not argue that shutdown or an added correction mechanism would fail.","attested":{"date":"2001-06-26","note":"The quoted wording also appears in the 2001 HTML edition as captured on 2001-06-26; the quotation itself is taken from MIRI’s later reprint."}}}},{"id":"yk-y24-time-2023-shut-it-all-down-c","statement":"“Pausing AI Developments Isn't Enough. We Need to Shut it All Down” (2023-03-29) contests Y24. Within his proposal for a worldwide moratorium, he rejects framing AI development as a national arms race on the ground that everyone shares the outcome. Discussing a moratorium that might begin in the United States, he describes the technology as one that no country could own and that would kill people in both countries.","evidenceIds":["evidence-yk-t24-a","evidence-yk-t24-b"]},{"id":"yk-y24-sl4-2007-11-24-yudkowsky-0711-17248-c","statement":"The SL4 message “Re: UCaRtMaAI paper” (2007-11-24) contests Y24. Replying to Tim Freeman, who argued against waiting until we know what we are doing because other AGI projects are proceeding without concern for Friendliness, he says that doing better than competitors does not count.","evidenceIds":["evidence-yk-t24-s4"]},{"id":"yk-y24-extropians-2000-09-29-yudkowsky-6080-a","statement":"The Extropians message “Re: Eugene's nuclear threat” (2000-09-29) bears on Y24. Correcting his earlier post in a thread on Eugene Leitl's suggestion of nuclear strikes on AI facilities, he says the only things he can do are to write up Friendly AI and to try to be the first to build AI.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: Racing to be first improves the outcome. Being first is tied to writing up Friendly AI, not to beating a named rival; the same reply rejects coercion against other projects. It does not state that whoever builds advanced AI controls it, or that racing to be first improves the outcome. Quotation kept short because the message has about 140 words of his own. His later texts on racing (SL4 message of 2007-11-24: building an AI is pass-fail; TIME, 2023: 'anyone talking of arms races is a fool') do not refer back to this message and do not supply the missing premise.","evidenceIds":["evidence-yk-t24-e2"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"Seeks to build AI first while writing up Friendly AI and rejecting coercion; it does not state that the builder controls the outcome or that racing to be first improves it."}}},{"id":"ec-m01-aghion-jones-jones-2017-ai-and-economic-growth-b","statement":"Aghion, Jones, Jones, “Artificial Intelligence and Economic Growth” (2017-10) bears on M01. Places AI in a continuous history of automation from the steam engine to semiconductors, while noting AI may now reach non-routine cognitive tasks.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"NBER Working Paper 23928, October 2017 (NBER Conference on AI paper; later a chapter in The Economics of Artificial Intelligence). Hedged (\"Perhaps\").","evidenceIds":["evidence-ec-m01-18"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"Not a pre-2015 text; a 2017 statement of continuity between AI and earlier automation.","tier":{"value":1,"basis":"NBER working paper (core venue)."}}}},{"id":"ec-m02-nourse-1955-automation-hearings-statement-b","statement":"Nourse, “Statement (former Chairman, Council of Economic Advisers), in Automation and Technological Change: Hearings” (1955-10-28) bears on M02. Speaker: Edwin G. Nourse, former Chairman, Council of Economic Advisers (economist). Nourse locates the 'real change' in electronic devices that remove the need for human judgement in process control.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The quote begins mid-sentence because 'pos¬sible' is broken by a non-standard OCR hyphen. Matches the cached .txt.","evidenceIds":["evidence-ec-m02-01"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"Describes machines taking over the mental element of control work, but not within a framework that allocates tasks between capital and labour, and not as AI.","tier":{"value":1,"basis":"Prepared statement by an economist (former Chairman of the Council of Economic Advisers, shown in the text); a substantive argued statement. Tier 1 by maintainer decision of 2026-10-12 (economist testimony)."},"conditions":"'to considerable extent'; limited to control of continuous production processes."}}},{"id":"lg-cn-yun-yeyi-minimax-coo-2025","statement":"MiniMax’s Hong Kong prospectus of 31 December 2025 identifies Yun Yeyi as executive Director and chief operating officer.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A role shown in the prospectus biography at the filing date. The record takes no start date for the COO title from other parts of the filing, and does not establish a current role.","evidenceIds":["lg-ev-cn-yun-yeyi-minimax-coo-2025"],"attributes":{"role":"executive Director and chief operating officer"}},{"id":"lg-oai-weil-departure-2026","statement":"Kevin Weil posted on X on 17 April 2026 that it was his last day at OpenAI; he stated that OpenAI for Science was being decentralized into other research teams.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A first-person announcement. The organisational change he mentions is his statement, not independently checked.","evidenceIds":["lg-ev-oai-weil-departure-2026"],"attributes":{"role":"Not stated as a title; he describes moving from Chief Product Officer to the research team and starting OpenAI for Science","statedReason":{"by":"kevin-weil","text":"Weil wrote that his last day came as OpenAI for Science was being decentralized into other research teams.","evidenceId":"lg-ev-oai-weil-departure-2026"}}},{"id":"lg-new-ito-sakana-cofounder","statement":"Sakana AI’s corporate information page names Ren Ito as a co-founder.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A company’s own founder list; it does not establish each person’s individual founding date.","evidenceIds":["lg-ev-new-ito-sakana-cofounder","lg-ev-new-ito-sakana-cofounder-b"]},{"id":"ap-pascual-restrepo-acemoglu-restrepo-2018-race-between-man-and-machine","statement":"“The Race between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment” (2018-06) credits Pascual Restrepo as one of its 2 authors.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Credited co-authorship; it does not allocate parts of the text between authors or imply agreement with every passage.","evidenceIds":["evidence-ap-pascual-restrepo-acemoglu-restrepo-2018-race-between-man-and-machine"]},{"id":"ap-eliezer-yudkowsky-extropians-1999-08-03-yudkowsky-1487","statement":"The mailing-list message “Re: IA vs. AI (vs. humanity)” (1999-08-03) 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-1999-08-03-yudkowsky-1487"]},{"id":"ap-eliezer-yudkowsky-lw-2021-ngo-and-yudkowsky-on-alignment-difficulty","statement":"“Ngo and Yudkowsky on alignment difficulty” (2021-11-15) credits Eliezer Yudkowsky as one of its 2 authors.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Credited co-authorship; it does not allocate parts of the text between authors or imply agreement with every passage.","evidenceIds":["evidence-ap-eliezer-yudkowsky-lw-2021-ngo-and-yudkowsky-on-alignment-difficulty"]},{"id":"cmp-rsp-v3-0-d1-cbrn","statement":"RSP · v3.0 states a provision on the comparison question “Chemical, biological, radiological or nuclear weapons”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Chemical and biological weapons are named in two threshold rows. Radiological and nuclear weapons are not named as thresholds; “CBRN” appears only in changelog entries about earlier versions. 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-rsp-v3-0-d1-cbrn-1","ev-cmp-rsp-v3-0-d1-cbrn-2"],"attributes":{"provision":{"kind":"risk_domain","key":"comparison-d1-cbrn","facet":{"id":"D1.cbrn","status":"stated"}}}},{"id":"cmp-fsf-2024-d5-cadence","statement":"FSF · 2024 states a provision on the comparison question “Evaluation timing”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"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-fsf-2024-d5-cadence-1"],"attributes":{"provision":{"kind":"evaluation_timing","key":"comparison-d5-cadence","facet":{"id":"D5.cadence","status":"stated"}}}},{"id":"c2-bytedance-seed-oss-licence","statement":"The Seed-OSS model card (revision of 26 August 2025) states that the Seed-OSS series is released under the Apache-2.0 license, and the seed-oss Git repository's LICENSE file at its commit of 25 August 2025 is the Apache License, Version 2.0.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Read from the Instruct model card and the code repository; the two Base repositories carry the apache-2.0 tag in the Hugging Face listing but their cards were not opened. 'Open-source' is Seed's term.","evidenceIds":["c2-ev-bytedance-seed-oss-licence","c2-ev-bytedance-seed-oss-licence-file"]},{"id":"fl-eu-aia-art51-classification-conditions","statement":"Article 51(1) provides that a general-purpose AI model is classified as a general-purpose AI model with systemic risk if it meets either of two conditions: (a) it has high impact capabilities evaluated with appropriate technical tools and methodologies, including indicators and benchmarks; or (b) a Commission decision finds it has equivalent capabilities or impact having regard to the criteria in Annex XIII.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"States what the Regulation requires of the categories it defines. Whether any named developer's model falls within a category, and whether any developer complies, is legal application and is not recorded. A statutory duty is not evidence of implementation or enforcement. Regulation (EU) 2026/1744, inspected 2026-10-27, does not list this article among its amendments (its Article 1 amends Articles 1-6, 10, 11, 17, 25, 27-30, 40, 42, 43, 50, 56(6), 57, 58, 60, 63, 64, 69, 70, 72, 75-77, 95-97, 99, 111, 113 and Annexes I and VIII). The compute presumption for point (a) is in recorded separately.","evidenceIds":["evidence-fl-eu-aia-art51-classification-conditions-1","evidence-fl-eu-aia-art51-classification-conditions-2"],"attributes":{"provision":{"kind":"threshold","key":"aia-art51-1"}}},{"id":"fl-eu-cop-final-changes-from-draft1-risk-types","statement":"The final Code's Appendix 1.4 specifies four systemic risks (CBRN, loss of control, cyber offence, harmful manipulation), where the initial draft's Measure 6.1 listed six types, including automated use of models for AI research and development and large-scale discrimination, which are not among the final four.","derivation":"inferred","resolution":"supported","inferenceRationale":"Both lists are quoted at their locators; the comparison is a set difference of the listed headings.","editorialState":"seed","qualification":"Compares the headings of the two lists. The two types are not deleted from the final Code: its description of loss of control includes autonomously creating or improving AI models (Appendix 1.4(2)); Appendix 1.3 lists capabilities to automate AI research and development as a source of systemic risk; Appendix 1.1 lists risks to non-discrimination among the types of risk to draw on. The record says only that the two types are not among the four specified risks. The intermediate drafts were not retrieved.","evidenceIds":["evidence-fl-eu-cop-final-changes-from-draft1-risk-types-1","evidence-fl-eu-cop-final-changes-from-draft1-risk-types-2","evidence-fl-eu-cop-final-changes-from-draft1-risk-types-3","evidence-fl-eu-cop-final-changes-from-draft1-risk-types-4"],"attributes":{"changeBasis":"document_comparison"}},{"id":"fl-intl-haip-first-round-listed","statement":"The OECD paper 'How are AI developers managing risks?' (September 2025) lists, in Annex A, 20 entities that submitted reports under the HAIP reporting framework between 7 February and 15 June 2025, including Anthropic, Google, Microsoft and OpenAI.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A list of report submitters, not of organisations found to apply the code. Google DeepMind is not listed as a separate submitter; 'Google' is. The object list is taken from the acknowledgements because the annex table extracts imperfectly (one row's name is missing in the text layer). Other official counts have different coverage and are not reconciled: the Japanese ministry's Hiroshima AI Process page says responses from 19 companies were published on the OECD website in April 2025, and the OECD.AI overview says the opening round resulted in 25 reports.","evidenceIds":["evidence-fl-intl-haip-first-round-listed-1","evidence-fl-intl-haip-first-round-listed-2","evidence-fl-intl-haip-first-round-listed-3","evidence-fl-intl-haip-first-round-listed-4"],"attributes":{"listing":{"count":20,"countBasis":"stated"}}},{"id":"cpo-orb-lumen-wp-published","statement":"Lumen Orbit (now Starcloud) published the white paper \"Why we should train AI in space\", whose version 1.01 is dated September 2024.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The date is the one the document prints. The earliest copy located is an Internet Archive capture of 24 December 2024, which the cross-family check found byte-identical to the copy retrieved in 2026 and which prints the same month; a capture does not establish when v1.01 was first posted. A later v1.03 carries the same month and adds the 2025 renaming.","evidenceIds":["evidence-cpo-orb-lumen-wp-published-1"]},{"id":"cpo-dc-meta-hyperion-5gw-2025","statement":"Mark Zuckerberg said on Meta's 30 July 2025 earnings call that Meta was building out Hyperion, \"which will be able to scale up to 5GW over several years\".","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Spoken remarks in a company transcript; a projection over unspecified years. The basis of the gigawatts is not stated.","evidenceIds":["evidence-cpo-dc-meta-hyperion-5gw-2025-1"],"attributes":{"transactionType":"compute_capacity","relationshipContext":"external","transactionStage":"planned","quantity":{"value":5,"unit":"GW (as worded: \"scale up to 5GW over several years\")","bound":"up_to","measures":"not_stated"}}},{"id":"cpo-dc-stargate-norway-230mw","statement":"The Stargate Norway release stated planned capacity of 230MW, with ambitions to expand by a further 290MW.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Planned figures; the 290MW is an ambition. What the megawatts measure is not stated.","evidenceIds":["evidence-cpo-dc-stargate-norway-230mw-1"],"attributes":{"transactionType":"compute_capacity","relationshipContext":"external","transactionStage":"planned","quantity":{"value":230,"unit":"MW (as worded: \"230MW of capacity\")","bound":"exact","measures":"not_stated"}}},{"id":"cpo-pw-amazon-constellation-calvert","statement":"Constellation and Amazon announced on 30 September 2026 a 20-year agreement covering 690 MW of power from the Calvert Cliffs plant in Maryland, including an uprate of approximately 190 MW expected online between 2030 and 2032, with a related retail supply agreement for Amazon operations in PJM.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Announced; the uprate is planned capacity. The release says Calvert Cliffs output continues to flow to the PJM grid; it does not name a data-centre site.","evidenceIds":["evidence-cpo-pw-amazon-constellation-calvert-1","evidence-cpo-pw-amazon-constellation-calvert-2","evidence-cpo-pw-amazon-constellation-calvert-3"],"attributes":{"transactionStage":"contracted","quantity":{"value":690,"unit":"megawatts of power, including a 190 MW uprate (as worded)","bound":"exact","measures":"contracted_energy"},"statedDates":[{"kind":"target_operation","wording":"2030–2032, uprate","statedBy":"constellation-energy"}],"power":{"technology":"nuclear","mechanism":"ppa"},"term":"20 years"}},{"id":"ab-arg-d1-sandbrink-articulates","statement":"Sandbrink's preprint on AI and biological misuse (arXiv v1, 24 June 2023) states that large language models such as GPT-4 are already able to provide dual-use information that could have enabled historical biological weapons efforts to succeed.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The record documents what version 1 of this preprint states; it is not evidence that the proposition is true, implies no influence between texts, and does not carry over to the later versions, which were not retrieved.","evidenceIds":["evidence-ab-arg-d1-sandbrink-articulates-1","evidence-ab-arg-d1-sandbrink-articulates-2"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","qualifier":"Stated as a claim about information provision, with historical programmes as the comparison."}}}],"sources":[{"id":"source-iem","title":"Intelligence Explosion Microeconomics","authors":["Eliezer Yudkowsky"],"publisher":"Machine Intelligence Research Institute","publicationDate":"2013","version":"Technical report 2013-1; last modified September 13, 2013","versionDate":"2013-09-13","canonicalUrl":"https://intelligence.org/files/IEM.pdf","evidenceUrl":"https://intelligence.org/files/IEM.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260917110159/https://intelligence.org/files/IEM.pdf","archivedAt":"2026-09-17","accessedAt":"2026-09-17","reuse":{"attribution":"Machine Intelligence Research Institute, Intelligence Explosion Microeconomics (2013).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and minimal attributed excerpts only. Recheck terms before wider redistribution."}},{"id":"source-rsp-3-0","title":"Responsible Scaling Policy — v3.0","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":"2026-02-24","version":"Responsible Scaling Policy v3.0; effective date 2026-02-24.","versionDate":"2026-02-24","canonicalUrl":"https://anthropic.com/responsible-scaling-policy/rsp-v3-0","evidenceUrl":"https://anthropic.com/responsible-scaling-policy/rsp-v3-0","language":"en","archivedUrl":"https://web.archive.org/web/20260224203046/https://www.anthropic.com/responsible-scaling-policy/rsp-v3-0","archivedAt":"2026-02-24","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v3.0.","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-fsf-1-0","title":"Frontier Safety Framework — v1.0","authors":["Google DeepMind"],"publisher":"Google DeepMind","publicationDate":"2024-05-17","version":"Frontier Safety Framework v1.0; publication date 2024-05-17.","versionDate":"2024-05-17","canonicalUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/introducing-the-frontier-safety-framework/fsf-technical-report.pdf","evidenceUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/introducing-the-frontier-safety-framework/fsf-technical-report.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20240517140902/https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/introducing-the-frontier-safety-framework/fsf-technical-report.pdf","archivedAt":"2024-05-17","accessedAt":"2026-09-18","reuse":{"attribution":"Google DeepMind, Frontier Safety Framework — v1.0.","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-anthropic-coi-2023-03-31","title":"Amended and Restated Certificate of Incorporation of Anthropic, PBC (filed 2023-03-31)","authors":["Anthropic, PBC"],"publisher":"Anthropic","publicationDate":"2023-03-31","version":"Filed 2023-03-31 according to the Division of Corporations stamp on page 1 (file number 4860621); 24-page scan. Delaware-stamped scan published by AI Lab Watch in a shared Google Drive folder (linked from its post of 12 June 2024, which says the filings were obtained from the State of Delaware); downloaded directly on 2026-10-04 and read from page images, as the PDF has no text layer. The Internet Archive holds no capture of this file and no new capture was requested. Authenticity rests on the filing stamp and the consistent recitals of later certificates; no copy was obtained from the Division of Corporations. Publisher as printed on the inspected expression: Anthropic, PBC (filed with the Delaware Division of Corporations).","canonicalUrl":"https://drive.google.com/file/d/1i8Itsv7v7t32lQgzNRqiiO9j5HqSMUVj/view","evidenceUrl":"https://drive.google.com/file/d/1i8Itsv7v7t32lQgzNRqiiO9j5HqSMUVj/view","language":"en","archivedUrl":"https://web.archive.org/web/20261029125140/https://drive.usercontent.google.com/download?id=1i8Itsv7v7t32lQgzNRqiiO9j5HqSMUVj&export=download&authuser=0","archivedAt":"2026-10-29","accessedAt":"2026-10-04","reuse":{"attribution":"Anthropic, PBC, amended and restated certificate of incorporation (filed 2023-03-31), copy published by AI Lab Watch.","licence":"Public filing; 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-anthropic-coi-2025-07-03","title":"Amended and Restated Certificate of Incorporation of Anthropic, PBC (filed 2025-07-03)","authors":["Anthropic, PBC"],"publisher":"Anthropic","publicationDate":"2025-07-03","version":"Filed 2025-07-03 according to the Division of Corporations stamp on page 1 (file number 4860621); 26-page scan. Delaware-stamped scan published by AI Lab Watch in a shared Google Drive folder (linked from its post of 12 June 2024, which says the filings were obtained from the State of Delaware); downloaded directly on 2026-10-04 and read from page images, as the PDF has no text layer. The Internet Archive holds no capture of this file and no new capture was requested. Authenticity rests on the filing stamp and the consistent recitals of later certificates; no copy was obtained from the Division of Corporations. Publisher as printed on the inspected expression: Anthropic, PBC (filed with the Delaware Division of Corporations).","canonicalUrl":"https://drive.google.com/file/d/1yeFKDaHN-3o7Zruvd-72De1H1K20Crsh/view","evidenceUrl":"https://drive.google.com/file/d/1yeFKDaHN-3o7Zruvd-72De1H1K20Crsh/view","language":"en","archivedUrl":"https://web.archive.org/web/20261029125714/https://drive.usercontent.google.com/download?id=1yeFKDaHN-3o7Zruvd-72De1H1K20Crsh&export=download","archivedAt":"2026-10-29","accessedAt":"2026-10-04","reuse":{"attribution":"Anthropic, PBC, amended and restated certificate of incorporation (filed 2025-07-03), copy published by AI Lab Watch.","licence":"Public filing; 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-anthropic-ltbt-2023","title":"The Long-Term Benefit Trust","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":"2023-09-19","version":"Web page dated Sep 19, 2023, inspected live 2026-10-04 at its current address; quotes also matched in the Internet Archive capture of 2023-09-25 at the original address. The live page adds an undated footnote on later changes in the Trust’s membership.","canonicalUrl":"https://www.anthropic.com/news/the-long-term-benefit-trust","evidenceUrl":"https://www.anthropic.com/news/the-long-term-benefit-trust","language":"en","archivedUrl":"https://web.archive.org/web/20230925170803/https://www.anthropic.com/index/the-long-term-benefit-trust","archivedAt":"2023-09-25","accessedAt":"2026-10-04","reuse":{"attribution":"Anthropic, The Long-Term Benefit Trust (19 September 2023).","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-eu-code-safety-security","title":"General-Purpose AI Code of Practice: Safety and Security chapter (final version, 10 July 2025)","authors":["Chairs and vice-chairs of the Code’s working groups (independent experts)"],"publisher":"European Commission","publicationDate":"2025-07-10","version":"PDF served by the Commission on 2026-10-04 (a re-rendered file of 43 pages); it has the same 15,392 words as the Internet Archive capture of 2025-07-10 (10:58 UTC, 40 pages), with a different reading order on the cover page. Every quote matched in both expressions. Publisher as printed on the inspected expression: European Commission (AI Office).","canonicalUrl":"https://ec.europa.eu/newsroom/dae/redirection/document/118119","evidenceUrl":"https://ec.europa.eu/newsroom/dae/redirection/document/118119","language":"en","archivedUrl":"https://web.archive.org/web/20250710105829/https://ec.europa.eu/newsroom/dae/redirection/document/118119","archivedAt":"2025-07-10","accessedAt":"2026-10-04","reuse":{"attribution":"General-Purpose AI Code of Practice, Safety and Security chapter (10 July 2025), published by the European Commission.","licence":"European Union; quoted briefly under the editorial policy","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-cotra-2021-cold-takes","title":"Why AI alignment could be hard with modern deep learning","authors":["Ajeya Cotra"],"publisher":"Cold Takes (Holden Karnofsky’s blog), guest post","publicationDate":"2021-09-21","version":"Current page, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search; the quote also matches the raw capture of 2021-09-21, re-verified for this collection.","canonicalUrl":"https://www.cold-takes.com/why-ai-alignment-could-be-hard-with-modern-deep-learning/","evidenceUrl":"https://www.cold-takes.com/why-ai-alignment-could-be-hard-with-modern-deep-learning/","language":"en","archivedUrl":"https://web.archive.org/web/20210921172314/https://www.cold-takes.com/why-ai-alignment-could-be-hard-with-modern-deep-learning/","archivedAt":"2021-09-21","accessedAt":"2026-10-06","reuse":{"attribution":"Ajeya Cotra, \"Why AI alignment could be hard with modern deep learning\", Cold Takes, 21 September 2021.","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-cfai-miri-reprint","title":"Creating Friendly AI 1.0 (MIRI reprint)","authors":["Eliezer Yudkowsky"],"publisher":"Singularity Institute for Artificial Intelligence","publicationDate":"2001-06-15","version":"Singularity Institute, MIRI reprint, retrieved 2026-10-06 (SHA-256 51f21b8b70c2b7324c20de85edea40b2ef224cadff56ef7e0a3aa2a08dba2752). Release date from the reprint; an open-commentary version was announced on SL4 on 2001-04-18, so some wording may be earlier than this date. Reprint wording is not authenticated for 2001 unless a capture was compared.","canonicalUrl":"https://intelligence.org/files/CFAI.pdf","evidenceUrl":"https://intelligence.org/files/CFAI.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20261001205112/https://intelligence.org/files/CFAI.pdf","archivedAt":"2026-10-01","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky, “Creating Friendly AI 1.0 (MIRI reprint)”, Singularity Institute, MIRI reprint, 2001-06-15.","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-cfai-2001-challenge-html","title":"Creating Friendly AI 1.0, section 2: Challenges of Friendly AI (2001 HTML edition)","authors":["Eliezer Yudkowsky"],"publisher":"Singularity Institute for Artificial Intelligence","publicationDate":"2001-06-15","version":"Raw Internet Archive capture of 2001-06-26, retrieved 2026-10-08 (SHA-256 e910a4f4aec9673a1dc69d006a40d44315acb6279f10fc9e6e3d79f5701dc603); used to date wording also quoted from MIRI’s reprint, not quoted itself.","canonicalUrl":"http://singinst.org/CFAI/challenge.html","evidenceUrl":"https://web.archive.org/web/20010626231122/http://singinst.org/CFAI/challenge.html","language":"en","archivedUrl":"https://web.archive.org/web/20010626231122/http://singinst.org/CFAI/challenge.html","archivedAt":"2001-06-26","accessedAt":"2026-10-08","reuse":{"attribution":"Eliezer Yudkowsky, Creating Friendly AI 1.0 (Singularity Institute for Artificial Intelligence, 2001).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and labelled paraphrases. Recheck terms before wider redistribution."}},{"id":"source-extropians-2000-09-29-yudkowsky-6080","title":"Re: Eugene's nuclear threat (Extropians mailing list)","authors":["Eliezer Yudkowsky"],"publisher":"Extropians mailing list","publicationDate":"2000-09-29","version":"Archive HTML as served, retrieved 2026-10-07 (SHA-256 50b829ded65f6c396627fc2b6659cdba6d9186084b37fac2897801f7ada758de); 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.3Q00/6080.html","evidenceUrl":"http://extropians.weidai.com/extropians.3Q00/6080.html","language":"en","archivedUrl":"https://web.archive.org/web/20261029180838/http://extropians.weidai.com/extropians.3Q00/6080.html","archivedAt":"2026-10-29","accessedAt":"2026-10-07","reuse":{"attribution":"Eliezer Yudkowsky, “Re: Eugene's nuclear threat”, Extropians mailing list, 2000-09-29.","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-1999-08-03-yudkowsky-1487","title":"Re: IA vs. AI (vs. humanity) (Extropians mailing list)","authors":["Eliezer Yudkowsky"],"publisher":"Extropians mailing list","publicationDate":"1999-08-03","version":"Archive HTML as served, retrieved 2026-10-07 (SHA-256 9234ee148973e0bf6feb236801e09f00eea8907cb862e91a55a64f3d6e2b08a8); 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.3Q99/1487.html","evidenceUrl":"http://extropians.weidai.com/extropians.3Q99/1487.html","language":"en","archivedUrl":"https://web.archive.org/web/20220811153857/https://extropians.weidai.com/extropians.3Q99/1487.html","archivedAt":"2022-08-11","accessedAt":"2026-10-07","reuse":{"attribution":"Eliezer Yudkowsky, “Re: IA vs. AI (vs. humanity)”, Extropians mailing list, 1999-08-03.","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-lw-2021-ngo-and-yudkowsky-on-alignment-difficulty","title":"Ngo and Yudkowsky on alignment difficulty","authors":["Eliezer Yudkowsky","Richard Ngo"],"publisher":"LessWrong","publicationDate":"2021-11-15","version":"Post body retrieved through the LessWrong GraphQL API on 2026-10-06 (SHA-256 e9897f317f4628db3254b0a109a0646bfa797f79757b76dbb1fba80ad80b06f5), content version 1.0.0.","canonicalUrl":"https://www.lesswrong.com/posts/7im8at9PmhbT4JHsW/ngo-and-yudkowsky-on-alignment-difficulty","evidenceUrl":"https://www.lesswrong.com/posts/7im8at9PmhbT4JHsW/ngo-and-yudkowsky-on-alignment-difficulty","language":"en","archivedUrl":"https://web.archive.org/web/20260721232039/https://www.lesswrong.com/posts/7im8at9PmhbT4JHsW/ngo-and-yudkowsky-on-alignment-difficulty","archivedAt":"2026-07-21","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky and Richard Ngo, “Ngo and Yudkowsky on alignment difficulty”, LessWrong, 2021-11-15.","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-eu-ai-act-2024","title":"Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act)","authors":["European Parliament and Council of the European Union"],"publisher":"Official Journal of the European Union (EUR-Lex)","publicationDate":"2024-07-12","version":"Official Journal L series edition of 12.7.2024, English HTML, retrieved 2026-10-10. The text of the Internet Archive capture of 2024-07-12 matched it apart from the numbering of footnote markers in the amending articles. Later amendments (Regulation (EU) 2026/1744) are not applied to this text.","canonicalUrl":"https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202401689","evidenceUrl":"https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202401689","language":"en","archivedUrl":"https://web.archive.org/web/20240712193938/https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202401689","archivedAt":"2024-07-12","accessedAt":"2026-10-10","reuse":{"attribution":"European Parliament and Council of the European Union, Regulation (EU) 2024/1689 (Artificial Intelligence Act), Official Journal L, 12.7.2024.","licence":"Official legal text; 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-aghion-jones-jones-2017-nber-capture","title":"Artificial Intelligence and Economic Growth","authors":["Philippe Aghion","Benjamin F. Jones","Charles I. Jones"],"publisher":"NBER Working Paper 23928","publicationDate":"2017-10","version":"Wayback capture of the NBER PDF (w23928); compared with the author-hosted Version 1.0 (web.stanford.edu/~chadj/AI.pdf, 2018 capture). Capture retrieved 2026-10-08 (SHA-256 a8e68273c738c80fe565afa60bc27ec5e8aaf3135011a0f760233896f1f2100c). Versions: NBER Working Paper 23928 (2017-10): checked; author copy (Stanford), version 1.0 (2017-10-10): compared; book chapter, University of Chicago Press (2019): not retrieved. Published version: Aghion, P., Jones, B. F., & Jones, C. I. (2019). Artificial intelligence and economic growth. In A. Agrawal, J. Gans, & A. Goldfarb (Eds.), The economics of artificial intelligence: An agenda (pp. 237–290). University of Chicago Press. https://doi.org/10.7208/chicago/9780226613475.003.0009 Working paper: NBER Working Paper No. 23928 (2017). https://doi.org/10.3386/w23928. Re-inspected: Full text. Current copy retrieved directly on 2026-10-09; no archive capture (Wayback unreachable), so earlier bytes are not authenticated. SHA-256 a8e68273c738c80fe565afa60bc27ec5e8aaf3135011a0f760233896f1f2100c. (Merged on 2026-10-29 with the record of the re-inspection, source-aghion-jones-jones-2017-nber, by consistency audit fix B4.)","canonicalUrl":"https://www.nber.org/system/files/working_papers/w23928/w23928.pdf","evidenceUrl":"https://web.archive.org/web/20210303021500/https://www.nber.org/system/files/working_papers/w23928/w23928.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20210303021500/https://www.nber.org/system/files/working_papers/w23928/w23928.pdf","archivedAt":"2021-03-03","accessedAt":"2026-10-09","reuse":{"attribution":"Philippe Aghion, Benjamin F. Jones, Charles I. Jones, “Artificial Intelligence and Economic Growth”, NBER Working Paper 23928, 2017-10.","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-acemoglu-restrepo-2018-race-between-man-and-machine","title":"The Race between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment","authors":["Daron Acemoglu","Pascual Restrepo"],"publisher":"American Economic Review 108(6), 2018, pp. 1488–1542","publicationDate":"2018-06","version":"Abstract page only. Current copy retrieved directly on 2026-10-09; no archive capture (Wayback unreachable), so earlier bytes are not authenticated. SHA-256 ec349902ff22b013972448a3868cda67441112008727c5df0772aee87639ae4b. Full text not retrieved; only the AEA abstract page. Acemoglu, D., & Restrepo, P. (2018). The race between man and machine: Implications of technology for growth, factor shares, and employment. American Economic Review, 108(6), 1488–1542. https://doi.org/10.1257/aer.20160696","canonicalUrl":"https://www.aeaweb.org/articles?id=10.1257/aer.20160696","evidenceUrl":"https://www.aeaweb.org/articles?id=10.1257/aer.20160696","language":"en","accessedAt":"2026-10-09","reuse":{"attribution":"Daron Acemoglu, Pascual Restrepo, “The Race between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment”, American Economic Review 108(6), 2018, pp. 1488–1542, 2018-06.","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-nourse-1955-automation-hearings-statement","title":"Statement (former Chairman, Council of Economic Advisers), in Automation and Technological Change: Hearings","authors":["Edwin G. Nourse"],"publisher":"U.S. Congress, Joint Committee on the Economic Report, Subcommittee on Economic Stabilization. (1955). Automation and technological change: Hearings … 84th Congress, 1st session, October 14–28, 1955. Washington: U.S. Government Printing Office, from p. 618","publicationDate":"1955-10-28","version":"Internet Archive OCR text (djvu.txt) of a scanned copy, retrieved 2026-10-09 from https://ia803102.us.archive.org/32/items/congress-automation-and-technological-change/congress-automation-and-technological-changedjvu.txt; OCR errors possible; quotes checked only against clean stretches. Current copy; no earlier capture compared. SHA-256 3962809fbb923be62bf986addb0014ee147934a8767475c1f7cfe051959bff90.","canonicalUrl":"https://archive.org/details/congress-automation-and-technological-change","evidenceUrl":"https://archive.org/details/congress-automation-and-technological-change","language":"en","accessedAt":"2026-10-09","reuse":{"attribution":"Edwin G. Nourse, “Statement (former Chairman, Council of Economic Advisers), in Automation and Technological Change: Hearings”, U.S. Congress, Joint Committee on the Economic Report, Subcommittee on Economic Stabilization. (1955). Automation and technological change: Hearings … 84th Congress, 1st session, October 14–28, 1955. Washington: U.S. Government Printing Office, from p. 618, 1955-10-28.","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-minimax-prospectus","title":"MiniMax Group Inc. — Global Offering","authors":["MiniMax Group Inc."],"publisher":"MiniMax","publicationDate":"2025-12-31","version":"Dated document, retrieved on 2026-10-15 (SHA-256 7853021e46da5714242f8d27984498cbfef3e260d5a7cbb8d4e6d62d83ab9df3) directly from the publisher. Publisher as printed on the inspected expression: MiniMax Group Inc. (via HKEXnews).","canonicalUrl":"https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1231/2025123100025.pdf","evidenceUrl":"https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1231/2025123100025.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260930092156/https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1231/2025123100025.pdf","archivedAt":"2026-09-30","accessedAt":"2026-10-15","reuse":{"attribution":"MiniMax Group Inc., “MiniMax Group Inc. — Global Offering”, 2025-12-31.","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-seed-oss-readme","title":"Seed-OSS-36B-Instruct model card (README.md, revision 497f1dca)","authors":["ByteDance Seed Team"],"publisher":"ByteDance Seed (Hugging Face organisation ByteDance-Seed)","publicationDate":"2025-08-26","version":"Dated page, retrieved on 2026-10-15 (SHA-256 a8b639d3654d3453bed9037cd1470ea6bc43289b1058a71840f8e11a99a6ca49) directly from the publisher.","canonicalUrl":"https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct/raw/497f1dca95ebdec98e41d517b9f060ee753c902f/README.md","evidenceUrl":"https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct/raw/497f1dca95ebdec98e41d517b9f060ee753c902f/README.md","language":"en","archivedUrl":"https://web.archive.org/web/20261028235553/https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct/raw/497f1dca95ebdec98e41d517b9f060ee753c902f/README.md","archivedAt":"2026-10-28","accessedAt":"2026-10-15","reuse":{"attribution":"ByteDance Seed Team, “Seed-OSS-36B-Instruct model card (README.md, revision 497f1dca)”, 2025-08-26.","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-sakana-info","title":"Corporate Info — Sakana AI (company page, undated)","authors":["Sakana AI"],"publisher":"Sakana AI","publicationDate":null,"version":"Undated page, retrieved on 2026-10-15 (SHA-256 3ed2c29c0aaf2091548efeed35708539f04e7a56e97aa9169447772ee4156a3b) directly from the publisher. Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://sakana.ai/company-info/","evidenceUrl":"https://sakana.ai/company-info/","language":"en","archivedUrl":"https://web.archive.org/web/20261007203956/https://sakana.ai/company-info/","archivedAt":"2026-10-07","accessedAt":"2026-10-15","reuse":{"attribution":"Sakana AI, “Corporate Info — Sakana AI (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-oai-weil-x-2026","title":"Post of 17 April 2026 on his last day at OpenAI","authors":["Kevin Weil"],"publisher":"X","publicationDate":"2026-04-17","version":"Dated post, retrieved on 2026-10-15 (SHA-256 09c80c554059542d85133527f2f70b06e1fb6b5bd34d4e3d3d79481d68172c83) through X’s own syndication endpoint for the post’s text, which truncates long notes. Read through X’s syndication endpoint (text only). Publication date is the post’s UTC timestamp.","canonicalUrl":"https://x.com/kevinweil/status/2045230426210648348","evidenceUrl":"https://x.com/kevinweil/status/2045230426210648348","language":"en","archivedUrl":"https://web.archive.org/web/20260417195826/https://twitter.com/kevinweil/status/2045230426210648348","archivedAt":"2026-04-17","accessedAt":"2026-10-15","reuse":{"attribution":"Kevin Weil, “Post of 17 April 2026 on his last day at OpenAI”, 2026-04-17.","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":"c2-src-seed-oss-gh-license","title":"ByteDance-Seed/seed-oss LICENSE","authors":["ByteDance Seed"],"publisher":"ByteDance-Seed/seed-oss Git repository (GitHub)","publicationDate":"2025-08-25","version":"Dated page, retrieved on 2026-10-27 (SHA-256 41fe2167e8a52d375dd36304db345312bd629addb2e109e47104a7d34a2cf116) directly from the publisher. Version: 3dc7757688bcfe3742e0ff9e393936ec5383a22b. LICENSE at commit 'Update README.md' of 25 August 2025; the file is present, unchanged, from the initial commit 79f8ca3 of 20 August 2025.","canonicalUrl":"https://raw.githubusercontent.com/ByteDance-Seed/seed-oss/3dc7757688bcfe3742e0ff9e393936ec5383a22b/LICENSE","evidenceUrl":"https://raw.githubusercontent.com/ByteDance-Seed/seed-oss/3dc7757688bcfe3742e0ff9e393936ec5383a22b/LICENSE","language":"en","archivedUrl":"https://web.archive.org/web/20261028224048/https://raw.githubusercontent.com/ByteDance-Seed/seed-oss/3dc7757688bcfe3742e0ff9e393936ec5383a22b/LICENSE","archivedAt":"2026-10-28","accessedAt":"2026-10-27","reuse":{"attribution":"ByteDance Seed, “ByteDance-Seed/seed-oss LICENSE”, 2025-08-25.","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-eu-code-draft1-2024","title":"First Draft General-Purpose AI Code of Practice","authors":["Chairs and Vice-Chairs of the Working Groups of the General-Purpose AI Code of Practice"],"publisher":"European Commission","publicationDate":"2024-11-14","version":"Retrieved on 2026-10-28 (SHA-256 e7522e16b0dd05f368592ff19cd2e4af1f1208be88015bfa0059cdb5f6a32354). The PDF served by the Commission's newsroom (document 109946), 36 pages, every page marked “DRAFT DOCUMENT”; its metadata gives creation on 14 November 2024, 10:55 UTC. Decision 5 of 2026-10-28: the Commission's own file is cited. The candidate pack first read the copy held by the Dutch Senate (Eerste Kamer der Staten-Generaal, https://www.eerstekamer.nl/9370000/1/j4nvi0xeni9vr2l_j9vvkyblr2b3ny8/vmidr7qz6qv4/f=/vmidr9wfudtt.pdf, SHA-256 605816c07298146ec981fb46929518ee09d3072ea49994d2b1b032366555ba87); the cross-family check of 2026-10-28 found the two files' extracted text identical on all 36 pages, although the raw files differ.","canonicalUrl":"https://ec.europa.eu/newsroom/dae/redirection/document/109946","evidenceUrl":"https://ec.europa.eu/newsroom/dae/redirection/document/109946","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"Chairs and Vice-Chairs of the Working Groups of the General-Purpose AI Code of Practice, “First Draft General-Purpose AI Code of Practice”, 2024-11-14.","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-mic-haip-reporting-page","title":"Reporting Framework | Hiroshima AI Process","authors":["Ministry of Internal Affairs and Communications, Japan"],"publisher":"Ministry of Internal Affairs and Communications, Japan (Hiroshima AI Process site)","publicationDate":null,"version":"Retrieved on 2026-10-27 (SHA-256 348bf2a18e4c8f013bb6cb7d44954aa55e2b1bce00968e7aac6efabeedb56bc4). English page as served on 2026-10-27 Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.soumu.go.jp/hiroshimaaiprocess/en/report.html","evidenceUrl":"https://www.soumu.go.jp/hiroshimaaiprocess/en/report.html","language":"en","accessedAt":"2026-10-27","reuse":{"attribution":"Ministry of Internal Affairs and Communications, Japan, “Reporting Framework | Hiroshima AI Process”.","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-oecd-haip-overview","title":"The Hiroshima AI Reporting Framework (OECD.AI overview)","authors":["OECD"],"publisher":"OECD","publicationDate":null,"version":"Retrieved on 2026-10-27 (SHA-256 1ca98b625a137cf9e897e7b41f0e8f928eec11150ca368e83b8dc0b1914f3dfd). OECD.AI page as served on 2026-10-27 (describes Version 2.0 and a 30 September 2026 submission cut-off) Its wording can change; earlier wording is not authenticated. Publisher as printed on the inspected expression: OECD.AI Policy Observatory.","canonicalUrl":"https://oecd.ai/en/transparency/overview","evidenceUrl":"https://oecd.ai/en/transparency/overview","language":"en","accessedAt":"2026-10-27","reuse":{"attribution":"OECD, “The Hiroshima AI Reporting Framework (OECD.AI overview)”.","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-oecd-haip-managing-risks-2025","title":"How are AI developers managing risks? Insights from responses to the reporting framework of the Hiroshima AI Process Code of Conduct (OECD Artificial Intelligence Papers, No. 45)","authors":["OECD"],"publisher":"OECD","publicationDate":"2025-09","version":"Retrieved on 2026-10-27 (SHA-256 2b49417beaded517b58e67b07bc76cbbee07db3dfecbdf83211306b46e9eb961). Cover: 'September 2025 No. 45'. Publisher as printed on the inspected expression: OECD Publishing.","canonicalUrl":"https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/09/how-are-ai-developers-managing-risks_fbaeb3ad/658c2ad6-en.pdf","evidenceUrl":"https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/09/how-are-ai-developers-managing-risks_fbaeb3ad/658c2ad6-en.pdf","language":"en","accessedAt":"2026-10-27","reuse":{"attribution":"OECD, “How are AI developers managing risks? Insights from responses to the reporting framework of the Hiroshima AI Process Code of Conduct (OECD Artificial Intelligence Papers, No. 45)”, 2025-09.","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":"source-cpo-orb-lumen-wp-gh","title":"Why we should train AI in space (White Paper v1.01)","authors":["Ezra Feilden","Adi Oltean","Philip Johnston"],"publisher":"Lumen Orbit, Inc. (now Starcloud), via lumenorbit.github.io","publicationDate":"2024-09","version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 90e1bde5b01437d7778e044f8ed95df1c49a63559fc1e35177c94d6443e384f7). PDF v1.01 as served at lumenorbit.github.io/wp.pdf on 2026-10-28; cover line \"White Paper v1.01, September 2024\". Starcloud's current link (starcloud.com/wp) serves v1.03 (starcloudinc.github.io/wp.pdf), which adds a note that Lumen Orbit was renamed Starcloud in 2025; v1.01 is quoted here. The served file cannot authenticate when the v1.01 bytes were first published. Retrieved again at admission on 2026-10-30 (SHA-256 90e1bde5b01437d7778e044f8ed95df1c49a63559fc1e35177c94d6443e384f7, the same bytes). The cross-family check (2026-10-30) opened the Internet Archive capture linked here and found it byte-identical to this retrieval; a capture shows only what the archive held on its date.","canonicalUrl":"https://lumenorbit.github.io/wp.pdf","evidenceUrl":"https://lumenorbit.github.io/wp.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20241224070032/https://lumenorbit.github.io/wp.pdf","archivedAt":"2024-12-24","accessedAt":"2026-10-28","reuse":{"attribution":"Ezra Feilden, Adi Oltean, Philip Johnston, “Why we should train AI in space (White Paper v1.01)”, 2024-09.","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":"source-cpo-dc-nscale-stargate-norway-2025","title":"Nscale, Aker and OpenAI to establish Stargate Norway: a 100,000 NVIDIA GPU AI Gigafactory powered by renewable energy in Northern Norway","authors":["Nscale"],"publisher":"Nscale","publicationDate":"2025-07-31","version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 28c39e2b4b5c53558183257759186845c04ebbacd3f3a56d2124fbf291e00497). Press release page as served on 2026-10-28; dateline \"Oslo, Norway – 31 July 2025\". OpenAI's post (openai.com) refuses direct retrieval. Retrieved again at admission on 2026-10-30 (SHA-256 28c39e2b4b5c53558183257759186845c04ebbacd3f3a56d2124fbf291e00497, the same bytes). Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.nscale.com/press-releases/stargate-norway-nscale-aker-openai","evidenceUrl":"https://www.nscale.com/press-releases/stargate-norway-nscale-aker-openai","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"Nscale, “Nscale, Aker and OpenAI to establish Stargate Norway: a 100,000 NVIDIA GPU AI Gigafactory powered by renewable energy in Northern Norway”, 2025-07-31.","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-dc-meta-q2-2025-call","title":"Meta Platforms, Inc. (META) Second Quarter 2025 Results Conference Call (transcript)","authors":["Meta Platforms, Inc. (investor relations transcript PDF hosted on q4cdn.com)"],"publisher":"Meta Platforms, Inc. (investor relations transcript PDF hosted on q4cdn.com)","publicationDate":"2025-07-30","version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 96d255e30c6e21fa34ac53b3ebd2736ba8f03fa52063d9f8c532746a87b041de). Transcript PDF as served on 2026-10-28; header \"July 30th, 2025\". A company-prepared transcript of spoken remarks, not an audited filing. Retrieved again at admission on 2026-10-30 (SHA-256 96d255e30c6e21fa34ac53b3ebd2736ba8f03fa52063d9f8c532746a87b041de, the same bytes).","canonicalUrl":"https://s21.q4cdn.com/399680738/files/doc_financials/2025/q2/META-Q2-2025-Earnings-Call-Transcript.pdf","evidenceUrl":"https://s21.q4cdn.com/399680738/files/doc_financials/2025/q2/META-Q2-2025-Earnings-Call-Transcript.pdf","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"Meta Platforms, Inc. (investor relations transcript PDF hosted on q4cdn.com), “Meta Platforms, Inc. (META) Second Quarter 2025 Results Conference Call (transcript)”, 2025-07-30.","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-pw-constellation-amazon-calvert-2026","title":"Constellation and Amazon Announce 20-Year Power Purchase Agreement Adding 190 Megawatts of Nuclear Capacity at Calvert Cliffs","authors":["Constellation Energy"],"publisher":"Constellation Energy","publicationDate":"2026-09-30","version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 453c0719cb30ee292b400645e9b07c6d383a3b3e0f956c5120eccd6bcd74a18c). Newsroom page as served on 2026-10-28; dateline LUSBY, MD – SEPT. 30, 2026. Retrieved again at admission on 2026-10-30 (SHA-256 453c0719cb30ee292b400645e9b07c6d383a3b3e0f956c5120eccd6bcd74a18c, the same bytes). Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.constellationenergy.com/news/2026/09/constellation-and-amazon-announce-20-year-power-purchase-agreement-at-calvert-cliffs.html","evidenceUrl":"https://www.constellationenergy.com/news/2026/09/constellation-and-amazon-announce-20-year-power-purchase-agreement-at-calvert-cliffs.html","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"Constellation Energy, “Constellation and Amazon Announce 20-Year Power Purchase Agreement Adding 190 Megawatts of Nuclear Capacity at Calvert Cliffs”, 2026-09-30.","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-sandbrink-2023","title":"Artificial intelligence and biological misuse: Differentiating risks of language models and biological design tools","authors":["arXiv (preprint, Jonas B. Sandbrink, Nuffield Department of Medicine, University of Oxford)"],"publisher":"arXiv (preprint, Jonas B. Sandbrink, Nuffield Department of Medicine, University of Oxford)","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 1ab85d8647f2d23079d35eb8162b70b3f970094106808ef4291fb89615db0afa). arXiv v1 of 24 June 2023 (the article was revised to v8 on 23 December 2023; this record cites v1 only) Retrieved again at admission on 2026-10-30 (SHA-256 1ab85d8647f2d23079d35eb8162b70b3f970094106808ef4291fb89615db0afa, the same bytes).","canonicalUrl":"https://arxiv.org/pdf/2306.13952v1","evidenceUrl":"https://arxiv.org/pdf/2306.13952v1","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"arXiv (preprint, Jonas B. Sandbrink, Nuffield Department of Medicine, University of Oxford), “Artificial intelligence and biological misuse: Differentiating risks of language models and biological design tools”.","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-arxiv-2306-13952-abs","title":"arXiv abstract page for 2306.13952v1 (submission history)","authors":["arXiv (Cornell University)"],"publisher":"arXiv (Cornell University)","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 f05b805a071dd4ad81ec84279296ed530c8a4b679f8d05ffa4760d24a79d8440). Abstract page as served on 2026-10-28; used only for the v1 submission date and the version history Retrieved again at admission on 2026-10-30 (SHA-256 f05b805a071dd4ad81ec84279296ed530c8a4b679f8d05ffa4760d24a79d8440, the same bytes). Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://arxiv.org/abs/2306.13952v1","evidenceUrl":"https://arxiv.org/abs/2306.13952v1","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"arXiv (Cornell University), “arXiv abstract page for 2306.13952v1 (submission history)”.","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-yudkowsky-iem-2013","sourceRole":"primary","support":"supports","locator":"PDF page 1, title and byline","passage":"Eliezer Yudkowsky","passageType":"quote","verification":"machine_checked","checkedAt":"2026-09-17","note":"Source passage inspected by an automated research assistant. Human editorial review is pending."},{"id":"evidence-cs-anthropic-ltbt-elects-directors-1","sourceRole":"primary","support":"supports","locator":"Article VIII(B)4(b) (page 20)","passage":"the Electing Class T Holders, exclusively and as a separate class, shall be entitled to elect three directors of the Corporation","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-04","note":"Verbatim quote transcribed from the page image by the integrating session on 2026-10-04 and compared with the scan; the copy has no reliable text layer, so the match is visual, not mechanical. The passage was first proposed and machine-verified in the week 4b research pass. Human editorial review is pending."},{"id":"evidence-cs-anthropic-ltbt-elects-directors-2","sourceRole":"primary","support":"supports","locator":"Article VIII(B)6(v) (page 21)","passage":"“First Phase-In Date” means September 6, 2023;","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-04","note":"Verbatim quote transcribed from the page image by the integrating session on 2026-10-04 and compared with the scan; the copy has no reliable text layer, so the match is visual, not mechanical. The passage was first proposed and machine-verified in the week 4b research pass. Human editorial review is pending."},{"id":"evidence-cs-anthropic-ltbt-elects-directors-3","sourceRole":"primary","support":"supports","locator":"Section “LTBT: Basic Structure and Features”, second paragraph","passage":"The Class T stock grants the Trust the authority to elect and remove a number of Anthropic’s board members that will phase in according to time- and funding-based milestones","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-04","note":"Verbatim quote, re-checked on 2026-10-04 by the integrating session against its own direct fetch of the cited expression (mechanical match after whitespace normalisation). Also matched in the Internet Archive capture of 2023-09-25. The passage was first proposed and machine-verified in the week 4b research pass. Human editorial review is pending."},{"id":"evidence-cs-anthropic-ltbt-elects-directors-4","sourceRole":"primary","support":"context","locator":"Article VIII(B)2–4 (page 22)","passage":"The 2025 certificate keeps the same allocation: one, then two, then three of five directors elected by the Electing Class T Holders as a separate class.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-04","note":"Editorial paraphrase of the page image, re-checked on 2026-10-04 by the integrating session against the scan. The integrating session wrote or selected this passage from its own reading of the source; it does not repeat a passage from the week 4b research pass. Human editorial review is pending."},{"id":"evidence-ts-b5-cotra-2021-context","sourceRole":"primary","support":"supports","locator":"Section on the analogy, paragraph mapping Saints, Sycophants and Schemers to models","passage":"a “Schemer” could be a model that seems to perform well because performing well during training will give it more opportunities to pursue its own goals later.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"The same passage as the B4 record."},{"id":"evidence-yk-t01-a","sourceRole":"primary","support":"supports","locator":"Creating Friendly AI 1.0, section 2 'Challenges of Friendly AI', paragraph beginning on printed p. 2 ('The term “Friendly AI” refers to…'); quoted sentence on p. 3","passage":"Similarly, it is “conservative” (see below) to assume that AIs cannot be forcibly constrained.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Candidate pack passage T01-a. The quotation (14 words) matched the retrieved expression (SHA-256 51f21b8b70c2b7324c20de85edea40b2ef224cadff56ef7e0a3aa2a08dba2752) when checked on 2026-10-06."},{"id":"evidence-yk-t01-a-2001","sourceRole":"primary","support":"supports","locator":"The 2001 HTML edition, page challenge.html","passage":"The same wording as the quotation from MIRI’s reprint appears in the 2001 HTML edition captured on 2001-06-26.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-08","note":"Compared on 2026-10-08 after normalising quotation marks, dashes and whitespace."},{"id":"evidence-yk-t24-a","passage":"anyone talking of arms races is a fool. That we all live or die as one, in this, is not a policy but a fact of nature.","passageType":"quote"},{"id":"evidence-yk-t24-s4","passage":"It doesn't matter if someone else does worse.","passageType":"quote"},{"id":"evidence-yk-t24-e2","sourceRole":"primary","support":"supports","locator":"message body, paragraph beginning 'The only thing I can…'","passage":"try to *be* the first, which I'm also doing","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-07","note":"Candidate pack passage T24-e2. The quotation (9 words) matched the retrieved expression (SHA-256 50b829ded65f6c396627fc2b6659cdba6d9186084b37fac2897801f7ada758de) when checked on 2026-10-07.","contextOnly":{"reason":"An earlier position later contradicted but not explicitly retracted; never show alone","requires":"His later texts on racing, which do not refer back to this message.","claimIds":["yk-y24-sl4-2007-11-24-yudkowsky-0711-17248-c","yk-y24-time-2023-shut-it-all-down-c"]}},{"id":"evidence-ec-m01-18","sourceRole":"primary","support":"supports","locator":"pdf page 5; section 2 A.I. and Automation of Production (opening)","passage":"Perhaps artificial intelligence is the next phase of this process rather than a discrete break.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-14","note":"Candidate pack passage M01-18 (research/economics-mainstream). The quotation (15 words) matched the retrieved current copy (SHA-256 a8e68273c738c80fe565afa60bc27ec5e8aaf3135011a0f760233896f1f2100c) in the pack's verification. Re-checked at admission on 2026-10-14 against a copy retrieved that day.","contextOnly":{"reason":"Post-2015 text adopting a continuity framing; relevant as context for how M01-type framing persisted, not as an M01 instance.","requires":"Pre-2015 tier-1 texts."}},{"id":"evidence-ec-m02-01","sourceRole":"primary","support":"supports","locator":"printed p. 619 (statement, 'Four aspects of automation')","passage":"to dispense to considerable extent with the mental element in manual control and to use the feedback principle extensively.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-14","note":"Candidate pack passage M02-01 (research/economics-mainstream). The quotation (19 words) matched the retrieved current copy (SHA-256 3962809fbb923be62bf986addb0014ee147934a8767475c1f7cfe051959bff90) in the pack's verification. Re-checked at admission on 2026-10-14 against a copy retrieved that day.","contextOnly":{"reason":"Fresh-context audit (A1) judged that the quotation misleads when shown alone.","requires":"The fragment lacks its subject: 'electronic devices make it possible to dispense...'. It refers to continuous-process production control, not AI."}},{"id":"lg-ev-cn-yun-yeyi-minimax-coo-2025","sourceRole":"primary","support":"supports","locator":"Directors and Senior Management, “Executive Directors”, biography of Yun Yeyi, printed page 352","passage":"is our executive Director and chief operating officer","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. 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-weil-departure-2026","sourceRole":"primary","support":"supports","locator":"Post text, first two sentences","passage":"Weil writes that it is his last day at OpenAI as OpenAI for Science is decentralized into other research teams, and that he went from Chief Product Officer to the research team and started OpenAI for Science.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “last day at OpenAI” 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. 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":"x_syndication"},{"id":"lg-ev-new-ito-sakana-cofounder","sourceRole":"primary","support":"supports","locator":"Section “Leadership”","passage":"Ren Ito Co-Founder & Chairman","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (5 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-new-ito-sakana-cofounder-b","sourceRole":"primary","support":"supports","locator":"Section “About”","passage":"founded by David Ha (CEO), Ren Ito (Chairman), and Llion Jones (CTO) in 2023","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (14 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-pascual-restrepo-acemoglu-restrepo-2018-race-between-man-and-machine","sourceRole":"primary","support":"supports","locator":"Page head: author byline","passage":"Pascual Restrepo","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":"evidence-ap-eliezer-yudkowsky-extropians-1999-08-03-yudkowsky-1487","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":"evidence-ap-eliezer-yudkowsky-lw-2021-ngo-and-yudkowsky-on-alignment-difficulty","sourceRole":"primary","support":"supports","locator":"Post author list in the publisher’s post data (LessWrong GraphQL API)","passage":"Eliezer 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 through the publisher’s post data. Human editorial review is pending.","retrievalRoute":"direct"},{"id":"ev-cmp-rsp-v3-0-d1-cbrn-1","sourceRole":"primary","support":"supports","locator":"PDF p. 6, §1 Our Recommendations for Industry-Wide Safety, threshold table, row title, left column ('Capability or usage threshold')","passage":"The edition’s threshold table has a row titled “Non-novel chemical/biological weapons production”.","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. The row concerns AI systems able to give significant help to people with a basic technical background (the example given is undergraduate STEM degrees) in making or obtaining and deploying such weapons. Human editorial review is pending."},{"id":"ev-cmp-rsp-v3-0-d1-cbrn-2","sourceRole":"primary","support":"supports","locator":"PDF p. 7, §1, threshold table, next row title, left column","passage":"The next row, “Novel chemical/biological weapons production”, concerns AI systems able to give significant help to threat actors in making or obtaining and deploying chemical or biological weapons whose potential damage would greatly exceed earlier catastrophes like COVID-19.","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. The edition’s example of such threat actors is teams with moderate resources and expert backing. Human editorial review is pending."},{"id":"ev-cmp-fsf-2024-d5-cadence-1","sourceRole":"primary","support":"supports","locator":"PDF p. 2 (printed p. 2), Framework, component '2 - Evaluating frontier models', second paragraph, first sentence","passage":"The edition says Google DeepMind aims to evaluate its models at every sixfold increase in effective compute and after every three months of fine-tuning progress.","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. Footnote 4 says these evaluations will be ones Google DeepMind is confident a model will pass while it is still a sixfold effective-compute step, or three months of fine-tuning, short of a CCL. Human editorial review is pending."},{"id":"c2-ev-bytedance-seed-oss-licence","sourceRole":"primary","support":"supports","locator":"Section 'Introduction'","passage":"We release this series of models to the open-source community under the Apache-2.0 license.","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 (14 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. A cross-family machine check (2026-10-27, item K045) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"c2-ev-bytedance-seed-oss-licence-file","sourceRole":"primary","support":"supports","locator":"LICENSE, heading","passage":"Apache License Version 2.0, January 2004","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 (6 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. A cross-family machine check (2026-10-27, item K045) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-eu-aia-art51-classification-conditions-1","sourceRole":"primary","support":"supports","locator":"Article 51(1), chapeau","passage":"shall be classified as a general-purpose AI model with systemic risk if it meets any of the following conditions","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (19 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 the Publications Office's Cellar repository (the same Official Journal expression as the dataset's EUR-Lex source; EUR-Lex returned empty responses). 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-fl-eu-aia-art51-classification-conditions-2","sourceRole":"primary","support":"supports","locator":"Article 51(1), point (a)","passage":"it has high impact capabilities evaluated on the basis of appropriate technical tools and methodologies, including indicators and benchmarks","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (19 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 the Publications Office's Cellar repository (the same Official Journal expression as the dataset's EUR-Lex source; EUR-Lex returned empty responses). 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-fl-eu-cop-final-changes-from-draft1-risk-types-1","sourceRole":"primary","support":"supports","locator":"Appendix 1.4, points (1) to (4) (p. 37)","passage":"(2) Loss of control","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (4 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 ec.europa.eu. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. Headings of points (1)-(4): Chemical, biological, radiological and nuclear; Loss of control; Cyber offence; Harmful manipulation. A cross-family machine check (2026-10-28, item 25) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-eu-cop-final-changes-from-draft1-risk-types-2","sourceRole":"primary","support":"supports","locator":"Appendix 1.4, point (4) (p. 37)","passage":"(4) Harmful manipulation","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (3 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 ec.europa.eu. 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 (2026-10-28, item 25) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-eu-cop-final-changes-from-draft1-risk-types-3","sourceRole":"primary","support":"supports","locator":"Measure 6.1, list (p. 17)","passage":"Automated use of models for AI Research and Development","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (9 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 of the Commission’s own file, which this record cites (SHA-256 e7522e16b0dd05f368592ff19cd2e4af1f1208be88015bfa0059cdb5f6a32354). Read directly from the Commission's newsroom. 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 (2026-10-28, item 25) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-eu-cop-final-changes-from-draft1-risk-types-4","sourceRole":"primary","support":"supports","locator":"Measure 6.1, list (p. 17)","passage":"Large-scale discrimination: Large-scale illegal discrimination of individuals, communities, or societies.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (10 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 of the Commission’s own file, which this record cites (SHA-256 e7522e16b0dd05f368592ff19cd2e4af1f1208be88015bfa0059cdb5f6a32354). Read directly from the Commission's newsroom. 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 (2026-10-28, item 25) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-intl-haip-first-round-listed-1","sourceRole":"primary","support":"supports","locator":"Annex A, PDF p. 31","passage":"This annex provides the names of the 20 entities that submitted reports between 7 February and 15 June 2025.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (19 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.oecd.org. 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 (2026-10-28, item 15) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-intl-haip-first-round-listed-2","sourceRole":"primary","support":"supports","locator":"Acknowledgements, PDF p. 5","passage":"AI21 Labs, Anthropic, Data Privacy and AI, Fayston Preparatory School, Fujitsu, Google, KDDI Corporation, KYP.ai GmbH, MGOIT, Microsoft, NEC Corporation, NTT Corporation, OpenAI, Palo Alto Research / West Lake Research & Education Service, Preferred Networks, Rakuten Group, Salesforce, SoftBank, TELUS and TELUS Digital","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (41 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.oecd.org. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. Introduced as 'organisations that responded to the Hiroshima AI Process reporting framework'. A cross-family machine check (2026-10-28, item 15) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-intl-haip-first-round-listed-3","sourceRole":"primary","support":"context","locator":"Main text","passage":"Paraphrase: the initial responses, from 19 companies, were published on the OECD website in April 2025.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The paraphrase is Selekoa’s, written from the text retrieved on 2026-10-27 and read directly from www.soumu.go.jp; it is not a quotation and was not machine-matched. A cross-family machine check (2026-10-28, item 15) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-intl-haip-first-round-listed-4","sourceRole":"primary","support":"context","locator":"Overview, first paragraph","passage":"The first round resulted in 25 reports","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (7 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 of the same address (the page’s bytes had changed). Read directly from oecd.ai. 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 (2026-10-28, item 15) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-orb-lumen-wp-published-1","sourceRole":"primary","support":"supports","locator":"PDF p. 1, cover line","passage":"© Lumen Orbit, Inc., White Paper v1.01, September 2024","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (8 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. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-dc-meta-hyperion-5gw-2025-1","sourceRole":"primary","support":"supports","locator":"PDF p. 2","passage":"We're also building out Hyperion, which will be able to scale up to 5GW over several years.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (17 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. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-dc-stargate-norway-230mw-1","sourceRole":"primary","support":"supports","locator":"paragraph 2","passage":"Stargate Norway is planned to deliver 230MW of capacity, with ambitions to expand by an additional 290MW","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (17 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. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-pw-amazon-constellation-calvert-1","sourceRole":"primary","support":"supports","locator":"paragraph beginning “The expansion of Calvert Cliffs' generating…”","passage":"a 20-year agreement, which includes 690 megawatts of power, including the 190 megawatt uprate","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). 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. A cross-family machine check (2026-10-29/30, item 31) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-pw-amazon-constellation-calvert-2","sourceRole":"primary","support":"supports","locator":"paragraph beginning “LUSBY, MD – SEPT. 30, 2026…”","passage":"approximately 190 megawatts in total of new, emissions-free generating capacity coming online between 2030 and 2032","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (16 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 31) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-pw-amazon-constellation-calvert-3","sourceRole":"primary","support":"supports","locator":"paragraph beginning “The expansion of Calvert Cliffs' generating…”","passage":"a related retail supply agreement to support Amazon operations in the 13-state PJM market","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). 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. A cross-family machine check (2026-10-29/30, item 31) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-arg-d1-sandbrink-articulates-1","sourceRole":"primary","support":"supports","locator":"Abstract, PDF p. 1","passage":"LLMs, such as GPT-4, are already able to provide dual-use information that could have enabled historical biological weapons efforts to succeed.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (21 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 text is the primary source for its own statement. A cross-family machine check (2026-10-30, item 99) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-arg-d1-sandbrink-articulates-2","sourceRole":"primary","support":"supports","locator":"Submission history block, v1 and v8 entries","passage":"[v1] Sat, 24 Jun 2023 … [v8] Sat, 23 Dec 2023","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). 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. Added after the audit of 2026-10-28: both dates in time.note rested on no cited passage. Cited for the submission dates only; the retrieved PDF of v1 carries no arXiv stamp. A cross-family machine check (2026-10-30, item 99) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"}],"at":{"entities":[0,1,5,19,22,23,45,157,214,248,253,311,332,345,350,372,392,400,414,491,513,514,527,558,567,643,720,776,785,836,912,996,1007,1028,1091,1113,1140,1157,1161,1176,1248,1254],"claims":[20,660,967,984,1143,1149,1150,1310,1315,1826,1865,2021,2308,2370,2461,2529,2596,2806,2866,2886,2959,3111,3134,3144,3169,3415],"sources":[10,39,44,222,225,226,242,337,346,347,359,367,381,457,478,511,520,582,620,729,769,823,846,860,861,862,920,967,975,1008,1065,1066],"evidence":[21,816,817,818,819,1190,1209,1210,1386,1394,1395,1639,1644,2345,2406,2612,2613,2955,3017,3108,3178,3179,3260,3574,3575,3673,3674,3709,3710,3711,3712,3820,3821,3822,3823,4098,4148,4171,4229,4230,4231,4704,4705]},"version":"1lg47t7"}