{"entities":[{"id":"openai","name":"OpenAI"},{"id":"anthropic","name":"Anthropic"},{"id":"google-deepmind","name":"Google DeepMind"},{"id":"eliezer-yudkowsky","name":"Eliezer Yudkowsky"},{"id":"rsp-2023","name":"Responsible Scaling Policy (2023)"},{"id":"fsf-2024","name":"Frontier Safety Framework (2024)"},{"id":"rsp-v2-1","name":"Anthropic Responsible Scaling Policy — v2.1"},{"id":"fsf-v2-0","name":"Google DeepMind Frontier Safety Framework — v2.0"},{"id":"faif-2025","name":"xAI Frontier Artificial Intelligence Framework — December 2025 edition"},{"id":"mira-murati","name":"Mira Murati"},{"id":"meta","name":"Meta"},{"id":"meta-open-strategy-2024","name":"Open Source AI Is the Path Forward (2024)"},{"id":"ijepa-research-account","name":"I-JEPA research account (2023)"},{"id":"self-supervised-learning","name":"Self-supervised learning"},{"id":"vas-narasimhan","name":"Vas Narasimhan"},{"id":"minimax","name":"MiniMax"},{"id":"wh-voluntary-ai-commitments-2023-07","name":"Ensuring Safe, Secure, and Trustworthy AI: voluntary AI commitments (July 2023)"},{"id":"turing-1950-computing-machinery","name":"Computing Machinery and Intelligence"},{"id":"yudkowsky-2008-hard-takeoff","name":"Hard Takeoff"},{"id":"agi-ruin-2022","name":"AGI Ruin: A List of Lethalities"},{"id":"observed-behaviour-as-evidence","name":"Observed behaviour under training or oversight as evidence"},{"id":"fix-problems-as-they-arise","name":"Problems can be fixed as they arise"},{"id":"extropians-2000-09-05-yudkowsky-4277","name":"Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...)"},{"id":"new-york-raise-act-chapter-96-2026","name":"New York Senate Bill S8828 / Assembly Bill A9449 (2026): RAISE Act chapter amendment, signed as Chapter 96"},{"id":"ai-developer-incentives-and-safety","name":"AI developers’ incentives matter for safety"},{"id":"hanson-2008-wrapping-up","name":"Wrapping Up"},{"id":"tencent","name":"Tencent / 腾讯"},{"id":"varun-mohan","name":"Varun Mohan"},{"id":"alan-turing","name":"Alan Turing"},{"id":"cn-ai-content-labelling-measures-2025","name":"Measures for Labelling Artificial Intelligence-Generated Synthetic Content (人工智能生成合成内容标识办法)"},{"id":"lumen-orbit-wp-2024","name":"Why we should train AI in space"},{"id":"oc2-radiative-cooling-advantage","name":"Radiative cooling in orbit as an advantage"},{"id":"uk-synthetic-nucleic-acids-screening-guidance-2024","name":"UK screening guidance on synthetic nucleic acids for users and providers"}],"claims":[{"id":"xm-open-strategy","statement":"Meta published Zuckerberg’s argument for an open-model ecosystem.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The essay states the company’s strategy. Its use of “open source” is not an independent finding that every model or licence meets an external definition.","evidenceIds":["evidence-xm-open-strategy"]},{"id":"xm-ijepa-self-supervision","statement":"The I-JEPA account discusses self-supervised learning from unlabeled observations.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A described learning approach, not proof of human-like intelligence, generality or priority.","evidenceIds":["evidence-xm-ijepa-self-supervision"]},{"id":"prov-rsp-v2-1-no-train-deploy-without-safeguards","statement":"RSP v2.1 states that its Responsible Scaling Policy, released in September 2023, is a public commitment not to train or deploy models capable of causing catastrophic harm unless Anthropic has implemented safety and security measures that will keep risks below acceptable levels.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"What the edition states, quoted from the cited expression. It is not evidence of implementation, compliance or effectiveness.","evidenceIds":["evidence-prov-rsp-v2-1-no-train-deploy-without-safeguards"],"attributes":{"provision":{"kind":"required_action","key":"rsp-no-train-deploy-without-safeguards"}}},{"id":"cs-narasimhan-anthropic-board-2026","statement":"Anthropic announced on 14 April 2026 that the Long-Term Benefit Trust had appointed Vas Narasimhan to its board of directors and that, with this appointment, Trust-appointed directors made up a majority of the board.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A documented board role and Anthropic’s own statement about the board’s composition. The post names six other directors (Dario Amodei, Daniela Amodei, Yasmin Razavi, Jay Kreps, Reed Hastings and Chris Liddell), more than the five-member board fixed by the latest inspected certificate; a later certificate was not inspected.","evidenceIds":["evidence-cs-narasimhan-anthropic-board-2026-1","evidence-cs-narasimhan-anthropic-board-2026-2"],"attributes":{"role":"Member of the board of directors (appointed by the Long-Term Benefit Trust)"}},{"id":"prov-wh-2023-07-duration","statement":"The July 2023 commitments text states that companies intend the voluntary commitments to remain in effect until regulations covering substantially the same issues come into force.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"What the text says about intended duration. Whether any regulation has ended the commitments, and their status after January 2025, need interpretation and are set aside.","evidenceIds":["evidence-prov-wh-2023-07-duration-1"],"attributes":{"provision":{"kind":"exception","key":"duration"}}},{"id":"ts-b5-agi-ruin","statement":"AGI Ruin says many alignment problems of superintelligence will not naturally appear at pre-dangerous, passively safe levels of capability, and that behavioural inspection cannot be relied on for facts an AI might want to deceive you about.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Items 13 and 31; item 13 gives the example of a system that deliberately looks more aligned to deceive its programmers.","evidenceIds":["evidence-ts-b5-agi-ruin","evidence-ts-b5-agi-ruin-2"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction"}}},{"id":"yk-y14-extropians-2000-09-05-yudkowsky-4277-c","statement":"The Extropians message “Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...)” (2000-09-05) contests Y14. Replying to a list member who expected oversight systems to destroy any sociopathic superintelligence, and such controls to be a research focus of the Singularity Institute, he says a superintelligence cannot be removed after the fact; the decisive choices have to be made before the seed AI becomes superintelligent.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: We get to learn from failures with powerful AI and try again, as with other technologies. The same reply goes on to say that this is not an area for research and that one should design an AI trusted to recover from the programmers' mistakes, a stance he later revised. Also bears on ‘Switching it off if it becomes dangerous’ (Y01).","evidenceIds":["evidence-yk-t14-e2"],"attributes":{"thought":{"roles":["contesting","earliest-located"],"register":"nonfiction","engagement":{"kind":"engages-premise","rationale":"Under the message-level rule, the passage, read with the paragraph, message or dialogue turn it belongs to, gives a reason against the premise that we get to learn from failures with powerful AI and try again, as with other technologies."}}}},{"id":"ny-ch96-civil-penalty","statement":"Section 1427(1) of the General Business Law, as added by the bill, provides that the attorney general may bring a civil action to recover a civil penalty of up to one million dollars for a first violation and up to three million dollars per subsequent violation, determined based on the severity of the violation, where a large frontier developer fails to publish or transmit a compliant required document, makes a prohibited statement, fails to report an incident as required, or fails to comply with its own frontier AI framework.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A ceiling and a power, not a record of any penalty. The bill also states that it creates no private right of action. It states what the text provides, not whether any named developer falls within it or complies. A document’s statement is not evidence of implementation, compliance or enforcement.","evidenceIds":["evidence-ny-ch96-civil-penalty-1"],"attributes":{"provision":{"kind":"threshold","key":"ny-civil-penalty-cap","parties":[{"role":"decides","actor":"Attorney general (civil action)"}]}}},{"id":"ec-e04-hanson-2008-wrapping-up-a","statement":"Hanson, “Wrapping Up” (2008-12-07), in “The Hanson-Yudkowsky AI-Foom Debate”, articulates E04. Hanson proposes liability and insurance for AI developers as the institutional answer to the friendliness problem.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Hanson's own policy proposal. \"Hurt\" is generic harm, not extinction specifically. Tier 2.","evidenceIds":["evidence-ec-e04-05"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","tier":{"value":2,"basis":"Blog debate (2008 Overcoming Bias posts) between an economist and a non-economist: tier 2 as an online debate; the maintainer confirmed tier 2 for the debate on 2026-10-12."}}}},{"id":"lg-oai-murati-cto-return-2023","statement":"Sam Altman’s message of 29 November 2023, published by OpenAI, stated that Mira Murati would return to her role as CTO.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"An announced return to a prior role; no end date is recorded here.","evidenceIds":["lg-ev-oai-murati-cto-return-2023"],"attributes":{"activityStage":"announced","role":"CTO"}},{"id":"lg-goo-mohan-gdm-2025","statement":"Demis Hassabis posted on X on 11 July 2025 that Google DeepMind was welcoming Windsurf founder Varun Mohan, with some of Windsurf's engineering team.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The post gives no title or start date. The Google–Windsurf agreement and his departure from Windsurf are recorded separately.","evidenceIds":["lg-ev-goo-varun-gdm-2025"],"attributes":{"role":"joined Google DeepMind (no title given)"}},{"id":"cmp-faif-2025-d4-determination","statement":"No passage meeting the comparison question “Determination of capability or safeguard adequacy” was found in FAIF · December 2025 edition within the inspected scope.","derivation":"inferred","resolution":"supported","inferenceRationale":"Premises: the comparison question asks for the following. The person, office or body the edition names as making or approving the determination that a threshold is reached, that safeguards are adequate, or that a model may be deployed. The inspected scope of FAIF · December 2025 edition was: Whole text layer, PDF pp. 1–11 (p. 12 blank): all sections read. Method: the edition was read in full and its text layer searched case-insensitively for decid/decision, approv, risk owner, leadership, officer, board, determin. Hits examined and not counted: PDF p. 9 '4. Governance Approach' designates 'risk owners' responsible for mitigating identified risks, periodic audits and monitoring for incidents, but does not give them (or any named person or body) the determination of thresholds, safeguard adequacy or deployment; PDF p. 9 'If we determine…' items refer to xAI generically during incident response; PDF p. 10 '5. Deployment Decisions' says 'We will also balance various factors when making deployment decisions' without naming a decision-maker. The conclusion is bounded by that scope and method.","editorialState":"seed","qualification":"A recorded absence within the stated scope: not a claim that the provision was deleted, and not a statement about practice or intent.","evidenceIds":["ev-cmp-faif-2025-d4-determination"],"attributes":{"provision":{"kind":"decision_authority","key":"comparison-d4-determination","facet":{"id":"D4.determination"}}}},{"id":"cmp-faif-2025-d4-oversight","statement":"No passage meeting the comparison question “Board or external oversight” was found in FAIF · December 2025 edition within the inspected scope.","derivation":"inferred","resolution":"supported","inferenceRationale":"Premises: the comparison question asks for the following. A role the edition gives to a board, board committee, trust, or external party in overseeing or reviewing those decisions. The inspected scope of FAIF · December 2025 edition was: Whole text layer, PDF pp. 1–10 (all sections read). Method: the edition was read in full and its text layer searched case-insensitively for board, council, third part, independent, external, oversight, audit, trust, review. Hits examined and not counted: No board or director is mentioned. PDF p. 8: 'we may also provide vetted and qualified external red teams or appropriate government agencies unredacted versions' concerns access to publications (recorded under “Transparency and reporting”), not review of decisions. PDF p. 9: risk owners are 'responsible for periodic audits to enforce framework implementation' (internal; recorded under “Internal reporting”). PDF p. 5: 'Independent third-party assessments' of cyber capability and PDF p. 3: 'trusted third-party safety auditors' as vetted users are evaluation inputs. No board or external party is given a role in reviewing threshold or deployment decisions. The conclusion is bounded by that scope and method.","editorialState":"seed","qualification":"A recorded absence within the stated scope: not a claim that the provision was deleted, and not a statement about practice or intent.","evidenceIds":["ev-cmp-faif-2025-d4-oversight"],"attributes":{"provision":{"kind":"decision_authority","key":"comparison-d4-oversight","facet":{"id":"D4.oversight"}}}},{"id":"ap-alan-turing-turing-1950-computing-machinery","statement":"“Computing Machinery and Intelligence” (1950-10) credits A. M. Turing as its author.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Credited authorship of this expression; it says nothing about priority for the ideas discussed.","evidenceIds":["evidence-ap-alan-turing-turing-1950-computing-machinery"]},{"id":"ap-eliezer-yudkowsky-yudkowsky-2008-hard-takeoff","statement":"“Hard Takeoff” (2008-12-02) credits Eliezer Yudkowsky as its author.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Credited authorship of this expression; it says nothing about priority for the ideas discussed.","evidenceIds":["evidence-ap-eliezer-yudkowsky-yudkowsky-2008-hard-takeoff"]},{"id":"cmp-rsp-2023-d7-external","statement":"RSP · 2023 states a provision on the comparison question “Public or external reporting”.","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-rsp-2023-d7-external-1"],"attributes":{"provision":{"kind":"reporting","key":"comparison-d7-external","facet":{"id":"D7.external","status":"stated"}}}},{"id":"cmp-fsf-2024-d1-autonomy","statement":"FSF · 2024 states a provision on the comparison question “Autonomy, self-improvement or AI research and development”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Two domains map here: “Autonomy” (Autonomy level 1, framed as misuse of a model with agency) and “Machine Learning R&D” (levels 1 and 2). Neither is framed as loss of control by the model itself; misaligned AI is named only in Future work (see “Other named domains”). 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-d1-autonomy-1","ev-cmp-fsf-2024-d1-autonomy-2"],"attributes":{"provision":{"kind":"risk_domain","key":"comparison-d1-autonomy","facet":{"id":"D1.autonomy","status":"stated"}}}},{"id":"cmp-fsf-v2-0-d7-internal","statement":"FSF · v2.0 states a provision on the comparison question “Internal reporting”, in a weaker form than the question’s criterion.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Stated, partly: the passage has what the question’s criterion asks for, but in a weaker form than the criterion’s own words. No separate internal reporting or record-keeping requirement is stated. Response plans go to corporate governance bodies for review and approval, and the same paragraph gives the Google DeepMind AGI Safety Council a periodic review of how the Framework is implemented (PDF p. 7). 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-v2-0-d7-internal-1"],"attributes":{"provision":{"kind":"reporting","key":"comparison-d7-internal","facet":{"id":"D7.internal","status":"partly"}}}},{"id":"c2-minimax-investor-tencent","statement":"MiniMax’s prospectus names Image Frame Investment (HK) Limited, which it describes as a wholly owned subsidiary of Tencent Holdings Limited, among its key Pre-IPO Investors, holding approximately 2.84% at the date of the listing application.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The prospectus does not, in the cited passage, attribute any round amount to this investor; no round total is attributed to it. Shareholding percentages are those the filing states for specified dates. No influence or control is inferred from the shareholding. The investing entity is the subsidiary; the parent is the subject as the filing identifies it.","evidenceIds":["c2-ev-minimax-investor-tencent"],"attributes":{"transactionType":"equity_investment","relationshipContext":"external","period":"Pre-IPO Investments (series not specified in the cited passage)","transactionStage":"invested"}},{"id":"fl-cn-labelling-explicit-implicit","statement":"Article 3 of the Labelling Measures divides labels for AI-generated synthetic content into explicit labels, presented in text, sound or graphics and clearly perceptible to users, and implicit labels, added to the content's file data by technical measures and not easily perceptible.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Definitions only.","evidenceIds":["evidence-fl-cn-labelling-explicit-implicit-1","evidence-fl-cn-labelling-explicit-implicit-2"],"attributes":{"provision":{"kind":"threshold","key":"cn-labelling-art3"}}},{"id":"cpo-orb-oc2-lumen-2024","statement":"The Lumen Orbit white paper (September 2024) argues that orbital data centres can lower cooling costs by passive radiative cooling to space.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Proponent; the same paper states the radiator requirement (recorded separately).","evidenceIds":["evidence-cpo-orb-oc2-lumen-2024-1"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","attested":{"date":"2024-12-24","note":"The cover prints September 2024 (v1.01). 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 for the pack and again at admission."}}}},{"id":"ab-fw-rsp-2023-asl3-redteam-cbrn-domains","statement":"RSP v1.0 states that the misuse domains covered by its ASL-3 deployment red-teaming measure should at a minimum include causes of extreme CBRN risks, and cybersecurity.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A stated requirement of the edition. It is not evidence that red-teaming was performed, nor that it would detect anything. Only the domain-scoping sentence of the red-teaming bullet is cited; the surrounding sentences describing elicitation were not recorded.","evidenceIds":["evidence-ab-fw-rsp-2023-asl3-redteam-cbrn-domains-1"],"attributes":{"provision":{"kind":"required_action","key":"rsp-asl3-redteam-domains"}}},{"id":"ab-scr-uk-guidance-addressees","statement":"The UK screening guidance states that it is intended to assist all entities involved in the provision and use of synthetic nucleic acids operating in the UK.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"States whom the guidance says it is intended to assist. It is not evidence that any entity followed it, and it does not establish a legal obligation or a territorial rule. Split under the one-claim-per-edge rule from recorded separately. The provision kind 'threshold' is the kind the dataset uses for a definition that decides who is covered; see openQuestions.","evidenceIds":["evidence-ab-scr-uk-guidance-addressees-1"],"attributes":{"provision":{"kind":"threshold","key":"uk-synthetic-nucleic-acids-guidance-addressees"}}}],"sources":[{"id":"source-rsp-1-0","title":"Responsible Scaling Policy — v1.0","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":"2023-09-19","version":"Responsible Scaling Policy v1.0; effective date 2023-09-19.","versionDate":"2023-09-19","canonicalUrl":"https://www-cdn.anthropic.com/1adf000c8f675958c2ee23805d91aaade1cd4613/responsible-scaling-policy.pdf","evidenceUrl":"https://www-cdn.anthropic.com/1adf000c8f675958c2ee23805d91aaade1cd4613/responsible-scaling-policy.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20230919161730/https://www-files.anthropic.com/production/files/responsible-scaling-policy-1.0.pdf","archivedAt":"2023-09-19","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — 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-rsp-2-1","title":"Responsible Scaling Policy — v2.1","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":null,"version":"Responsible Scaling Policy v2.1; effective date 2025-03-31.","versionDate":"2025-03-31","canonicalUrl":"https://www-cdn.anthropic.com/17310f6d70ae5627f55313ed067afc1a762a4068.pdf","evidenceUrl":"https://www-cdn.anthropic.com/17310f6d70ae5627f55313ed067afc1a762a4068.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20250402134435/https://www-cdn.anthropic.com/17310f6d70ae5627f55313ed067afc1a762a4068.pdf","archivedAt":"2025-04-02","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v2.1.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-fsf-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-fsf-2-0","title":"Frontier Safety Framework — v2.0","authors":["Google DeepMind"],"publisher":"Google DeepMind","publicationDate":"2025-02-04","version":"Frontier Safety Framework v2.0; publication date 2025-02-04.","versionDate":"2025-02-04","canonicalUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/updating-the-frontier-safety-framework/Frontier%20Safety%20Framework%202.0.pdf","evidenceUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/updating-the-frontier-safety-framework/Frontier%20Safety%20Framework%202.0.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20250205011727/https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/updating-the-frontier-safety-framework/Frontier%20Safety%20Framework%202.0.pdf","archivedAt":"2025-02-05","accessedAt":"2026-09-18","reuse":{"attribution":"Google DeepMind, Frontier Safety Framework — v2.0.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-faif-2025","title":"Frontier Artificial Intelligence Framework — December 2025 edition","authors":["xAI"],"publisher":"xAI","publicationDate":null,"version":"Frontier Artificial Intelligence Framework December 2025 edition; revision date 2025-12-30.","versionDate":"2025-12-30","canonicalUrl":"https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf","evidenceUrl":"https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260102171656/https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf","archivedAt":"2026-01-02","accessedAt":"2026-09-18","reuse":{"attribution":"xAI, Frontier Artificial Intelligence Framework — December 2025 edition.","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-atlas-openai-return","title":"Sam Altman returns as CEO, OpenAI has a new initial board","authors":["Sam Altman","Bret Taylor","OpenAI"],"publisher":"OpenAI","publicationDate":"2023-11-29","version":"November 29, 2023 announcement with a separately labelled December 8 update; current HTML inspected 2026-09-28. Re-inspected: Web page dated November 29, 2023, inspected live 2026-10-04; quotes also matched in the Internet Archive capture of 2023-11-30 (02:05 UTC). (Merged on 2026-10-29 with the record of the re-inspection, source-openai-new-initial-board-2023, by consistency audit fix B4.)","canonicalUrl":"https://openai.com/index/sam-altman-returns-as-ceo-openai-has-a-new-initial-board/","evidenceUrl":"https://openai.com/index/sam-altman-returns-as-ceo-openai-has-a-new-initial-board/","language":"en","archivedUrl":"https://web.archive.org/web/20231130020514/https://openai.com/blog/sam-altman-returns-as-ceo-openai-has-a-new-initial-board","archivedAt":"2023-11-30","accessedAt":"2026-10-04","reuse":{"attribution":"OpenAI. Sam Altman returns as CEO, OpenAI has a new initial board.","licence":"No redistribution licence asserted.","note":"Short attributed paraphrases only; original publication is linked, not redistributed."}},{"id":"source-xm-open-strategy","title":"Open Source AI Is the Path Forward","authors":["Mark Zuckerberg"],"publisher":"Meta","publicationDate":"2024-07-23","version":"July 23, 2024 essay with January 27, 2025 update shown.","canonicalUrl":"https://about.fb.com/news/2024/07/open-source-ai-is-the-path-forward/","evidenceUrl":"https://about.fb.com/news/2024/07/open-source-ai-is-the-path-forward/","language":"en","archivedUrl":"https://web.archive.org/web/20260923150356/https://about.fb.com/news/2024/07/open-source-ai-is-the-path-forward/","archivedAt":"2026-09-23","accessedAt":"2026-09-29","reuse":{"attribution":"Open Source AI Is the Path Forward — Meta","licence":"No redistribution licence asserted.","note":"Source-linked metadata and brief evidence paraphrases; original text remains with its publisher."}},{"id":"source-xm-ijepa","title":"I-JEPA: The first AI model based on Yann LeCun’s vision for more human-like AI","authors":["Meta"],"publisher":"Meta","publicationDate":"2023-06-13","version":"Dated publisher webpage as retrieved; not an authenticated historical capture.","canonicalUrl":"https://ai.meta.com/blog/yann-lecun-ai-model-i-jepa/","evidenceUrl":"https://ai.meta.com/blog/yann-lecun-ai-model-i-jepa/","language":"en","archivedUrl":"https://web.archive.org/web/20260726170343/https://ai.meta.com/blog/yann-lecun-ai-model-i-jepa/","archivedAt":"2026-07-26","accessedAt":"2026-09-29","reuse":{"attribution":"I-JEPA: The first AI model based on Yann LeCun’s vision for more human-like AI — Meta","licence":"No redistribution licence asserted.","note":"Source-linked metadata and brief evidence paraphrases; original text remains with its publisher."}},{"id":"source-anthropic-narasimhan-2026","title":"Long-Term Benefit Trust appoints Vas Narasimhan to Board of Directors","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":"2026-04-14","version":"Web page dated Apr 14, 2026, inspected live 2026-10-04; the board-majority sentence also matched in the Internet Archive capture of 2026-04-14 (17:37 UTC).","canonicalUrl":"https://www.anthropic.com/news/narasimhan-board","evidenceUrl":"https://www.anthropic.com/news/narasimhan-board","language":"en","archivedUrl":"https://web.archive.org/web/20260414173745/https://www.anthropic.com/news/narasimhan-board","archivedAt":"2026-04-14","accessedAt":"2026-10-04","reuse":{"attribution":"Anthropic, Long-Term Benefit Trust appoints Vas Narasimhan to Board of Directors (14 April 2026).","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-wh-commitments-2023-07","title":"Ensuring Safe, Secure, and Trustworthy AI (voluntary AI commitments, July 2023)","authors":["The White House"],"publisher":"The White House","publicationDate":"2023-07-21","version":"PDF linked from the fact sheet of 21 July 2023, inspected 2026-10-04 at its archived-site address; byte-identical (SHA-1) to the Internet Archive captures of 2023-07-21 at the original whitehouse.gov address. The Biden–Harris White House pages are now served from the archived site bidenwhitehouse.archives.gov.","canonicalUrl":"https://bidenwhitehouse.archives.gov/wp-content/uploads/2023/07/Ensuring-Safe-Secure-and-Trustworthy-AI.pdf","evidenceUrl":"https://bidenwhitehouse.archives.gov/wp-content/uploads/2023/07/Ensuring-Safe-Secure-and-Trustworthy-AI.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20230721165220/https://www.whitehouse.gov/wp-content/uploads/2023/07/Ensuring-Safe-Secure-and-Trustworthy-AI.pdf","archivedAt":"2023-07-21","accessedAt":"2026-10-04","reuse":{"attribution":"The White House, Ensuring Safe, Secure, and Trustworthy AI (July 2023).","licence":"US government work; 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-turing-1950-amt-b-19","title":"Computing Machinery and Intelligence (offprint, AMT/B/19)","authors":["A. M. Turing"],"publisher":"Mind","publicationDate":"1950-10","version":"Original Mind offprint, scanned by the archive; printed pp. 448–449 read on the page images for this collection (scans 16–17). Offprint held by the Turing Digital Archive, King’s College, Cambridge.","canonicalUrl":"https://turingarchive.kings.cam.ac.uk/publications-lectures-and-talks-amtb/amt-b-19","evidenceUrl":"https://turingarchive.kings.cam.ac.uk/publications-lectures-and-talks-amtb/amt-b-19","language":"en","archivedUrl":"https://web.archive.org/web/20261030045116/https://mediadepot.kings.cam.ac.uk/getmedia/81/be4a3d910d5d58e3911315113c60a09f6293fa896a5e3795998fa6b4fb72b71446336e3c7c92d5e08fd0fb2762b6f663279ba799d7fd562480b4bf63c63e02b9","archivedAt":"2026-10-30","accessedAt":"2026-10-06","reuse":{"attribution":"A. M. Turing, \"Computing Machinery and Intelligence\", Mind 59 (236), October 1950; offprint in the Turing Digital Archive, AMT/B/19.","licence":"No open redistribution licence asserted; the archive hosts the offprint with Oxford University Press’s permission. Page images are not reproduced.","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-yudkowsky-2008-hard-takeoff","title":"Hard Takeoff","authors":["Eliezer Yudkowsky"],"publisher":"Overcoming Bias","publicationDate":"2008-12-02","version":"Text checked in the post’s first stored version through the LessWrong GraphQL API on 2026-10-06; content version 1.0.0. Now hosted on LessWrong.","canonicalUrl":"https://www.lesswrong.com/posts/tjH8XPxAnr6JRbh7k/hard-takeoff","evidenceUrl":"https://www.lesswrong.com/posts/tjH8XPxAnr6JRbh7k/hard-takeoff","language":"en","archivedUrl":"https://web.archive.org/web/20181213042057/https://www.lesswrong.com/posts/tjH8XPxAnr6JRbh7k/hard-takeoff","archivedAt":"2018-12-13","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky, \"Hard Takeoff\", Overcoming Bias (now hosted on LessWrong), 2008-12-02.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-agi-ruin-2022","title":"AGI Ruin: A List of Lethalities","authors":["Eliezer Yudkowsky"],"publisher":"LessWrong and Alignment Forum","publicationDate":"2022-06-05","version":"Text checked in the post’s first stored version through the LessWrong GraphQL API on 2026-10-06.","canonicalUrl":"https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","evidenceUrl":"https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","language":"en","archivedUrl":"https://web.archive.org/web/20220605225854/https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","archivedAt":"2022-06-05","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky, \"AGI Ruin: A List of Lethalities\", LessWrong and Alignment Forum, 2022-06-05.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-extropians-2000-09-05-yudkowsky-4277","title":"Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...) (Extropians mailing list)","authors":["Eliezer Yudkowsky"],"publisher":"Extropians mailing list","publicationDate":"2000-09-05","version":"Archive HTML as served, retrieved 2026-10-07 (SHA-256 817049e5c2b33859a370b003534043a70de9c896f3dd32f7fbb9dfe948097f0b); 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.3Q00/4277.html","evidenceUrl":"http://extropians.weidai.com/extropians.3Q00/4277.html","language":"en","archivedUrl":"https://web.archive.org/web/20160414051453/http://extropians.weidai.com/extropians.3Q00/4277.html","archivedAt":"2016-04-14","accessedAt":"2026-10-07","reuse":{"attribution":"Eliezer Yudkowsky, “Re: Why would AI want to be friendly? (Was: Congratulations to Eli,Brian ...)”, Extropians mailing list, 2000-09-05.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-ny-s8828-assembly-record","title":"New York S08828 (Gounardes) / A09449 (Bores): summary, actions and text of the bill (New York State Assembly bill search)","authors":["New York State Legislature"],"publisher":"New York State Legislature","publicationDate":"2026-03-27","version":"Bill-search page for S8828 retrieved 2026-10-11. It shows the summary, the action list ending 03/27/2026 SIGNED CHAP.96, and the bill as printed on 2026-01-08; the page for A09449 shows the same text and actions. The chapter as printed in the session laws was not inspected. No archive capture is cited: the Internet Archive refused lookup requests (HTTP 429) on 2026-10-11, so none was checked. Publisher as printed on the inspected expression: New York State Assembly (bill search and legislative information).","canonicalUrl":"https://nyassembly.gov/leg/?default_fld=&leg_video=&bn=S8828&term=2025&Summary=Y&Actions=Y&Text=Y","evidenceUrl":"https://nyassembly.gov/leg/?default_fld=&leg_video=&bn=S8828&term=2025&Summary=Y&Actions=Y&Text=Y","language":"en","accessedAt":"2026-10-11","reuse":{"attribution":"New York State Legislature, S08828 / A09449 (2026), as published on the New York State Assembly bill-search site.","licence":"Official document; 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-hanson-2008-wrapping-up-miri-2013","title":"Wrapping Up","authors":["Robin Hanson"],"publisher":"Machine Intelligence Research Institute, The Hanson-Yudkowsky AI-Foom Debate (2013)","publicationDate":"2013","version":"Chapter 46 of The Hanson-Yudkowsky AI-Foom Debate (Machine Intelligence Research Institute, 2013), which compiles the 2008 Overcoming Bias debate with later posts; the compilation heads the post with its date. Current MIRI PDF, retrieved 2026-10-11. SHA-256 6b6dfd0416402e1776d7ddcece95298df2df0ed9bc0bd4a6f4bd81c4b1e5102a.","canonicalUrl":"https://intelligence.org/files/AIFoomDebate.pdf","evidenceUrl":"https://intelligence.org/files/AIFoomDebate.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260929131419/https://intelligence.org/files/AIFoomDebate.pdf","archivedAt":"2026-09-29","accessedAt":"2026-10-11","reuse":{"attribution":"Robin Hanson, “Wrapping Up”, in Machine Intelligence Research Institute, The Hanson-Yudkowsky AI-Foom Debate (2013).","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-goo-x-hassabis-windsurf-2025","title":"Post by Demis Hassabis welcoming Windsurf founders to Google DeepMind (11 July 2025)","authors":["Demis Hassabis"],"publisher":"X","publicationDate":"2025-07-11","version":"Dated post, retrieved on 2026-10-15 (SHA-256 29092581140fe9be474178c1f4cec345a9da0694fd70dcf126710fd3a4d941b8) through X’s own syndication endpoint for the post’s text, which truncates long notes.","canonicalUrl":"https://twitter.com/demishassabis/status/1943794383536652309","evidenceUrl":"https://twitter.com/demishassabis/status/1943794383536652309","language":"en","archivedUrl":"https://web.archive.org/web/20250711220749/https://twitter.com/demishassabis/status/1943794383536652309","archivedAt":"2025-07-11","accessedAt":"2026-10-15","reuse":{"attribution":"Demis Hassabis, “Post by Demis Hassabis welcoming Windsurf founders to Google DeepMind (11 July 2025)”, 2025-07-11.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-cn-labelling-measures-2025","title":"关于印发《人工智能生成合成内容标识办法》的通知 (Notice on issuing the Measures for Labelling AI-Generated Synthetic Content), 国信办通字〔2025〕2号","authors":["国家互联网信息办公室 (Cyberspace Administration of China)","工业和信息化部 (Ministry of Industry and Information Technology)","公安部 (Ministry of Public Security)","国家广播电视总局 (National Radio and Television Administration)"],"publisher":"中国政府网 (www.gov.cn), 国务院政策文件库","publicationDate":"2025-03-07","version":"Retrieved on 2026-10-27 (SHA-256 0c0d1d32be9f69843088b0ad2664734b401485166bd04bdfdf35d42d9b77b84a). gov.cn policy-library page (source given as 国家网信办网站; 成文日期 2025-03-07), retrieved 2026-10-27. The English in parentheses in the title is a machine translation.","canonicalUrl":"https://www.gov.cn/zhengce/zhengceku/202503/content_7014286.htm","evidenceUrl":"https://www.gov.cn/zhengce/zhengceku/202503/content_7014286.htm","language":"zh-Hans","accessedAt":"2026-10-27","reuse":{"attribution":"国家互联网信息办公室 (Cyberspace Administration of China), 工业和信息化部 (Ministry of Industry and Information Technology), 公安部 (Ministry of Public Security), 国家广播电视总局 (National Radio and Television Administration), “关于印发《人工智能生成合成内容标识办法》的通知 (Notice on issuing the Measures for Labelling AI-Generated Synthetic Content), 国信办通字〔2025〕2号”, 2025-03-07.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"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-ab-uk-synna-text","title":"UK screening guidance on synthetic nucleic acids for users and providers","authors":["GOV.UK (Department for Science, Innovation and Technology)"],"publisher":"GOV.UK (Department for Science, Innovation and Technology)","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 d12116062f1c3dbb8a384593927c08cf40ee37289dc82cf6011540b63bcb7dd8). The HTML attachment as served on 2026-10-28. Its screening methodology sections and its list of indicators of a suspicious transaction were read but are not cited. Retrieved again at admission on 2026-10-30 (SHA-256 52c2f4a8bfc22b7819ee11216d08966df73f980a87b187b9c4b7bd868a544262).","canonicalUrl":"https://www.gov.uk/government/publications/uk-screening-guidance-on-synthetic-nucleic-acids/uk-screening-guidance-on-synthetic-nucleic-acids-for-users-and-providers","evidenceUrl":"https://www.gov.uk/government/publications/uk-screening-guidance-on-synthetic-nucleic-acids/uk-screening-guidance-on-synthetic-nucleic-acids-for-users-and-providers","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"GOV.UK (Department for Science, Innovation and Technology), “UK screening guidance on synthetic nucleic acids for users and providers”.","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-xm-open-strategy","sourceRole":"primary","support":"supports","locator":"Essay opening and “Why open source AI is good for Meta” section.","passage":"The essay argues that an open ecosystem benefits Meta and developers.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-29","note":"Passage checked by machine against the identified source expression. No human editorial acceptance or independent verification of implementation is implied."},{"id":"evidence-xm-ijepa-self-supervision","sourceRole":"primary","support":"supports","locator":"“Capturing common-sense knowledge with self-supervised learning” section.","passage":"The account describes learning from unlabeled observations and predicting representations.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-29","note":"Passage checked by machine against the identified source expression. No human editorial acceptance or independent verification of implementation is implied."},{"id":"evidence-prov-rsp-v2-1-no-train-deploy-without-safeguards","sourceRole":"primary","support":"supports","locator":"PDF p. 2 (unnumbered), Executive Summary, first paragraph, first sentence","passage":"In September 2023, we released our Responsible Scaling Policy (RSP), a public commitment not to train or deploy models capable of causing catastrophic harm unless we have implemented safety and security measures that will keep risks below acceptable levels.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2025-04-02. Extraction note: Repeated in the Introduction (PDF p. 5, printed p. 1). Operative forms: §4.3 final sentence (PDF p. 14) and §6.2 'Monitoring pretraining' (PDF p. 15). Human editorial review is pending."},{"id":"evidence-cs-narasimhan-anthropic-board-2026-1","sourceRole":"primary","support":"supports","locator":"First paragraph","passage":"Vas Narasimhan has been appointed to Anthropic's Board of Directors by the Anthropic Long-Term Benefit Trust.","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). The passage was first proposed and machine-verified in the week 4b research pass. Human editorial review is pending."},{"id":"evidence-cs-narasimhan-anthropic-board-2026-2","sourceRole":"primary","support":"supports","locator":"Third paragraph","passage":"With Narasimhan's appointment, Trust-appointed directors now make up a majority of the Board.","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). The passage was first proposed and machine-verified in the week 4b research pass. Human editorial review is pending."},{"id":"evidence-prov-wh-2023-07-duration-1","sourceRole":"primary","support":"supports","locator":"Page 2, paragraph introducing the commitments","passage":"Companies intend these voluntary commitments to remain in effect until regulations covering substantially the same issues come into force.","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). The passage was first proposed and machine-verified in the week 4b research pass. Human editorial review is pending."},{"id":"evidence-ts-b5-agi-ruin","sourceRole":"primary","support":"supports","locator":"Section B.1 \"The distributional leap\", item 13","passage":"Many alignment problems of superintelligence will not naturally appear at pre-dangerous, passively-safe levels of capability.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Checked against the first stored version on 2026-10-06."},{"id":"evidence-ts-b5-agi-ruin-2","sourceRole":"primary","support":"supports","locator":"Section B.3, item 31","passage":"A strategically aware intelligence can choose its visible outputs to have the consequence of deceiving you, including about … whether the intelligence has acquired strategic awareness; you can't rely on behavioral inspection to determine facts about an AI which that AI might want to deceive you about.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"The same passage as the B2 record."},{"id":"evidence-yk-t14-e2","sourceRole":"primary","support":"supports","locator":"message body, paragraph beginning 'You must be joking.'","passage":"You cannot \"put down\" a superintelligence like some kind of wounded pet. The time for such decisions is before the seed AI reaches superintelligence, not after.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-07","note":"Candidate pack passage T14-e2. The quotation (26 words) matched the retrieved expression (SHA-256 817049e5c2b33859a370b003534043a70de9c896f3dd32f7fbb9dfe948097f0b) when checked on 2026-10-07."},{"id":"evidence-ny-ch96-civil-penalty-1","sourceRole":"primary","support":"supports","locator":"Section 1427, subdivision 1","passage":"The attorney general may bring a civil action to recover a civil penalty in an amount not to exceed one million dollars for a first violation and in an amount not to exceed three million dollars per subsequent violation","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-11","note":"Verbatim quote, matched mechanically (after whitespace normalisation) by the proposing machine agent against the text extracted from the file retrieved on 2026-10-11. Checked by fresh-context machine audit on 2026-10-11 (review/frontier-governance-2-audit-2026-10-11.md). Human editorial review is pending."},{"id":"evidence-ec-e04-05","sourceRole":"primary","support":"supports","locator":"\"Wrapping Up\", 7 December 2008, MIRI compilation (2013) ch. 46, post body; pdf p. 364","passage":"developers should have to show an ability, perhaps via insurance, to pay plausible hurt amounts if their creations hurt others.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-16","note":"Candidate pack passage E04-05 (research/economics-mainstream/esoteric). The quotation (20 words) matched the retrieved current copy (SHA-256 6b6dfd0416402e1776d7ddcece95298df2df0ed9bc0bd4a6f4bd81c4b1e5102a) in the pack's verification of 2026-10-15. Re-checked at admission on 2026-10-16 against a copy retrieved that day."},{"id":"lg-ev-oai-murati-cto-return-2023","sourceRole":"primary","support":"supports","locator":"Altman’s message, opening paragraph","passage":"Mira will return to her role as CTO.","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 through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-16/17) re-read the record’s sources and judged it partly supported; its findings were applied at admission. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-goo-varun-gdm-2025","sourceRole":"primary","support":"supports","locator":"Post text, first sentence","passage":"Hassabis welcomes Windsurf founders Varun Mohan and Douglas Chen and some of the Windsurf engineering team to Google DeepMind.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “founders @_mohansolo & Douglas Chen” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through X’s own syndication endpoint for the post’s text, which truncates long notes. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"x_syndication"},{"id":"ev-cmp-faif-2025-d4-determination","sourceRole":"primary","support":"supports","locator":"FAIF · December 2025 edition: Whole text layer, PDF pp. 1–11 (p. 12 blank): all sections read","passage":"Within the inspected scope, no passage meeting the comparison question “Determination of capability or safeguard adequacy” was found.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-26","note":"Absence recorded by an automated research assistant on 2026-10-15 for the framework comparison, audited in a fresh context, checked by a reviewer from a different model family on 2026-10-15 and re-checked on 2026-10-26. Human editorial review is pending."},{"id":"ev-cmp-faif-2025-d4-oversight","sourceRole":"primary","support":"supports","locator":"FAIF · December 2025 edition: Whole text layer, PDF pp. 1–10 (all sections read).","passage":"Within the inspected scope, no passage meeting the comparison question “Board or external oversight” was found.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-26","note":"Absence recorded by an automated research assistant on 2026-10-15 for the framework comparison, audited in a fresh context, checked by a reviewer from a different model family on 2026-10-15 and re-checked on 2026-10-26. Human editorial review is pending."},{"id":"evidence-ap-alan-turing-turing-1950-computing-machinery","sourceRole":"primary","support":"supports","locator":"Offprint PDF page 1 (printed page 433), “By” line beneath the title; surname in small capitals","passage":"A. M. TURING","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Passage and locator as inspected by a reviewer from a different model family on 2026-10-27 in a direct retrieval of the offprint embedded in the King’s College archive page; the scan has no text layer, so the page was checked visually. The preparer had matched the archive site’s heading “Alan Turing”, not the offprint’s byline. Human editorial review is pending.","retrievalRoute":"direct"},{"id":"evidence-ap-eliezer-yudkowsky-yudkowsky-2008-hard-takeoff","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-2023-d7-external-1","sourceRole":"primary","support":"supports","locator":"PDF p. 10 (printed p. 10), Procedural Commitments, numbered item 5","passage":"The edition lists among its procedural commitments making evaluation results public after a model is deployed, where possible: sometimes in the first model card, otherwise after a delay if that serves a broad safety interest.","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. ASL-3 deployment measures add a process for disclosing vulnerabilities and incidents to other labs, within security or legal limits (PDF p. 9); ASL-2 measures include publishing model cards (PDF p. 6). Human editorial review is pending."},{"id":"ev-cmp-fsf-2024-d1-autonomy-1","sourceRole":"primary","support":"supports","locator":"PDF p. 1 (printed p. 1), opening section, second paragraph, third sentence","passage":"The edition says its first CCLs cover four domains: Autonomy, Biosecurity, Cybersecurity and “Machine Learning R&D”.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-27","note":"Explicit paraphrase, written on 2026-10-27 by an automated research assistant after re-reading the passage in the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. The comparison quoted the passage; it is given here as a paraphrase to keep the edition within the quotation limit. Human editorial review is pending."},{"id":"ev-cmp-fsf-2024-d1-autonomy-2","sourceRole":"primary","support":"supports","locator":"PDF p. 2 (printed p. 2), footnote 2","passage":"A footnote defines “Autonomy” as the possible misuse of models with substantial capacity for agency and flexible use of actions and tools across long time horizons and several domains.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-27","note":"Explicit paraphrase, written on 2026-10-27 by an automated research assistant after re-reading the passage in the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. The comparison quoted the passage; it is given here as a paraphrase to keep the edition within the quotation limit. Human editorial review is pending."},{"id":"ev-cmp-fsf-v2-0-d7-internal-1","sourceRole":"primary","support":"supports","locator":"PDF p. 7 (printed p. 7), Governance and Accountability, 'Governance', first sentence","passage":"The edition says that, for Google models reaching alert thresholds, appropriate corporate governance bodies will review and approve the response plan; the “Google DeepMind AGI Safety Council” is one example given.","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 sentence also names the “Google DeepMind Responsibility and Safety Council” and the “Google Trust & Compliance Council”. Human editorial review is pending."},{"id":"c2-ev-minimax-investor-tencent","sourceRole":"primary","support":"supports","locator":"Same section, Tencent","passage":"is a wholly owned subsidiary of Tencent Holdings Limited (“Tencent”)","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 (10 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 K056) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-cn-labelling-explicit-implicit-1","sourceRole":"primary","support":"supports","locator":"Article 3 (第三条), second to fourth paragraphs","passage":"Labels comprise explicit labels (added in the content or the interaction interface, presented as text, sound, graphics and the like, clearly perceptible to users) and implicit labels (added to file data by technical measures, not easily perceptible to users).","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.gov.cn; it is not a quotation and was not machine-matched. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-cn-labelling-explicit-implicit-2","sourceRole":"primary","support":"supports","locator":"Article 3 (第三条), second paragraph","passage":"人工智能生成合成内容标识包括显式标识和隐式标识","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Added at the admission of the frameworks-to-law pack (research/frameworks-to-law). Added at admission: the schema requires a provision record to quote the edition, and the pack gave this provision by paraphrase only. The passage was read from a fresh retrieval at admission whose SHA-256 equals the pack’s, and is within the source’s quotation limit. English machine translation, not verified by a person: “labels for AI-generated synthetic content comprise explicit labels and implicit labels”. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-orb-oc2-lumen-2024-1","sourceRole":"primary","support":"supports","locator":"PDF p. 2, \"Why Data Centers in Space?\"","passage":"Orbital data centers can leverage lower cooling costs using passive radiative cooling in space to directly achieve low coolant temperatures.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (20 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 83) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-fw-rsp-2023-asl3-redteam-cbrn-domains-1","sourceRole":"primary","support":"supports","locator":"PDF p. 8 (printed p. 8), “ASL-3 Deployment Measures”, bullet “Successfully pass red-teaming”, second sentence","passage":"Misuse domains should at a minimum include causes of extreme CBRN risks, and cybersecurity.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (14 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-scr-uk-guidance-addressees-1","sourceRole":"primary","support":"supports","locator":"Guidance recommendations, third paragraph","passage":"This gene synthesis guidance is intended to assist all entities involved in the provision and use of synthetic nucleic acids operating in the UK.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (24 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 (SHA-256 52c2f4a8bfc22b7819ee11216d08966df73f980a87b187b9c4b7bd868a544262; the bytes differ from the pack’s retrieval). Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. The next sentence says the UK is also working internationally to promote screening practices. A cross-family machine check (2026-10-30, item 69) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"}],"at":{"entities":[0,1,2,19,21,22,43,50,56,124,157,167,169,171,221,248,250,258,317,333,345,363,389,501,588,593,809,814,931,1009,1091,1116,1210],"claims":[346,354,407,668,689,969,1083,1236,1534,1669,1794,2140,2141,2337,2447,2491,2591,2608,2825,2908,3082,3247,3308],"sources":[35,37,44,45,51,125,155,159,229,235,258,322,338,356,471,541,582,647,837,920,1053],"evidence":[369,377,439,834,835,866,1192,1193,1319,1545,1917,2103,2307,2777,2778,2984,3094,3138,3254,3255,3274,3605,3753,3754,4048,4459,4539]},"version":"1lg47t7"}