{"entities":[{"id":"openai","name":"OpenAI"},{"id":"anthropic","name":"Anthropic"},{"id":"xai","name":"xAI"},{"id":"microsoft","name":"Microsoft"},{"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":"rsp-v3-4","name":"Anthropic Responsible Scaling Policy — v3.4"},{"id":"fsf-v2-0","name":"Google DeepMind Frontier Safety Framework — v2.0"},{"id":"greg-brockman","name":"Greg Brockman"},{"id":"jimmy-ba","name":"Jimmy Ba"},{"id":"ppo-paper","name":"Proximal Policy Optimization Algorithms"},{"id":"summarize-feedback-paper","name":"Learning to summarize from human feedback"},{"id":"metr","name":"Model Evaluation & Threat Research"},{"id":"gpt4-system-card-2023","name":"GPT-4 System Card"},{"id":"metr-gpt4o-report","name":"METR preliminary GPT-4o evaluation (2024)"},{"id":"zhipu","name":"Zhipu AI / 智谱"},{"id":"meta-aasf-v2-0","name":"Meta Advanced AI Scaling Framework — Version 2"},{"id":"openai-inc","name":"OpenAI, Inc. (OpenAI Foundation)"},{"id":"wh-voluntary-ai-commitments-2023-07","name":"Ensuring Safe, Secure, and Trustworthy AI: voluntary AI commitments (July 2023)"},{"id":"moravec-1979-todays-computers","name":"Today’s Computers, Intelligent Machines and Our Future"},{"id":"franzen-1992-real-life-turing-test-reply","name":"Re: Real-life Turing test (Franzen)"},{"id":"ai-automated-ai-research","name":"AI-automated AI research"},{"id":"behavioural-tests-as-evidence","name":"Behavioural tests as evidence of what a system is"},{"id":"build-superintelligence-fast","name":"Build superintelligence as fast as possible"},{"id":"lw-2008-the-magnitude-of-his-own-folly","name":"The Magnitude of His Own Folly"},{"id":"eu-ai-act-2024","name":"Regulation (EU) 2024/1689 (Artificial Intelligence Act), as published in the Official Journal"},{"id":"cn-generative-ai-interim-measures-2023","name":"Interim Measures for the Management of Generative Artificial Intelligence Services (生成式人工智能服务管理暂行办法), Order No. 15"},{"id":"no-current-sign-of-explosive-growth","name":"Current data show no approach to explosive growth"},{"id":"machines-cheaper-at-every-task","name":"Machines cheaper than labour at every task"},{"id":"davidson-2021-explosive-economic-growth","name":"Could Advanced AI Drive Explosive Economic Growth?"},{"id":"eloundou-et-al-2023-gpts-are-gpts","name":"GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models"},{"id":"technology-and-the-american-economy-1966","name":"Technology and the American Economy: Report of the National Commission on Technology, Automation, and Economic Progress, Volume I"},{"id":"hanson-2011-debating-yudkowsky","name":"Debating Yudkowsky"},{"id":"baidu","name":"Baidu / 百度"},{"id":"graham-moreno","name":"Graham Moreno"},{"id":"liu-debing","name":"Liu Debing"},{"id":"sarah-heck","name":"Sarah Heck"},{"id":"shanghai-ai-lab","name":"Shanghai Artificial Intelligence Laboratory / 上海人工智能实验室"},{"id":"sun-tianxiang","name":"Sun Tianxiang"},{"id":"windsurf","name":"Windsurf"},{"id":"tyna-eloundou","name":"Tyna Eloundou"},{"id":"robin-hanson","name":"Robin Hanson"},{"id":"sail-frontier-ai-risk-management-framework-v1-0","name":"Frontier AI Risk Management Framework (v1.0)"},{"id":"cn-cybersecurity-law-amendment-2025","name":"Decision of the NPC Standing Committee on Amending the Cybersecurity Law of the People's Republic of China (全国人民代表大会常务委员会关于修改《中华人民共和国网络安全法》的决定)"},{"id":"new-delhi-declaration-2026","name":"AI Impact Summit Declaration (New Delhi Declaration on AI Impact), February 2026"},{"id":"nscale","name":"Nscale"},{"id":"stargate-norway","name":"Stargate Norway, Kvandal (Narvik), Norway"},{"id":"anthropic-llms-biorisk-2025-09","name":"LLMs and biorisk"},{"id":"ab-prop-evaluation-validity","name":"Published evaluations do not establish real-world biological risk"}],"claims":[{"id":"microsoft-investment-2019","statement":"Microsoft announced a $1 billion investment in OpenAI.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Announcement amount; cash disbursement, instrument and ownership share are not established.","evidenceIds":["evidence-microsoft-investment-2019"],"attributes":{"transactionType":"investment","relationshipContext":"external","transactionStage":"announced","amount":{"value":1000000000,"currency":"USD","bound":"exact","scope":"transaction"}}},{"id":"jimmy-ba-xai-team-2023","statement":"Jimmy Ba appeared on xAI’s July 2023 launch team roster.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Contemporaneous secondary report; original archive unavailable. Roster membership does not establish a co-founder title, start date or current role.","evidenceIds":["evidence-jimmy-ba-xai-team-2023"],"attributes":{"role":"Launch team member"}},{"id":"lineage-summaries-cites-ppo","statement":"Learning to summarize from human feedback cites PPO for policy optimisation.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"An explicit work-level reference only; neither agreement nor a causal influence claim. The citing text does not identify a particular arXiv revision of the target.","evidenceIds":["evidence-lineage-summaries-cites-ppo"]},{"id":"eval-i-metr-gpt4o-report","statement":"METR published its preliminary GPT-4o evaluation report.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The report calls the research informal and identifies elicitation and human-baseline limitations. Publication is not certification.","evidenceIds":["evidence-eval-i-metr-gpt4o-report"]},{"id":"prov-meta-aasf-v2-0-critical-threshold","statement":"Advanced AI Scaling Framework v2 states that a model meets the critical risk threshold if its continued development could substantially contribute to any threat scenario, or if it could uniquely enable a threat scenario and the risk cannot be mitigated in the proposed deployment context.","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-meta-aasf-v2-0-critical-threshold"],"attributes":{"provision":{"kind":"threshold","key":"meta-critical-threshold"}}},{"id":"provchange-fsf-v2-0-response-plan","statement":"v2.0 states that on reaching an alert threshold for a CCL the model's proximity to the CCL is assessed and the risk analysed, involving internal and external experts as needed, to inform the formulation and application of a response plan; v1.0 stated that on reaching evaluation thresholds a response plan would be formulated based on the analysis of the CCL and evaluation results.","derivation":"inferred","resolution":"supported","inferenceRationale":"Premises: FSF · v2.0 states the provision as quoted in its provision record (PDF p. 3 (printed p. 3), component “3 - Applying Mitigations”, first paragraph, first two sentences); FSF · 2024 states it as quoted in its provision record (PDF p. 2, component “3 - Applying mitigations”, first sentence). The difference described is read directly from the two passages.","editorialState":"seed","qualification":"A textual difference between the two editions’ wording of this keyed provision. It is not an assessment of whether the change strengthened or weakened the framework, and not evidence of implementation.","evidenceIds":["evidence-provchange-fsf-v2-0-response-plan-later","evidence-provchange-fsf-v2-0-response-plan-earlier"],"attributes":{"changeBasis":"document_comparison","provision":{"kind":"required_action","key":"fsf-response-plan"}}},{"id":"cm-link-rsp-2023-wh","statement":"Anthropic’s Responsible Scaling Policy v1.0 states that its ASL-2 deployment measures overlap substantially with the White House voluntary commitments that Anthropic and other companies made in July, which it says it continues to maintain.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The edition’s own words, recorded as naming the commitments: it describes overlap, not fulfilment. It is not evidence that the commitments are met.","evidenceIds":["evidence-cm-link-rsp-2023-wh-1"],"attributes":{"commitment":{"link":"names"}}},{"id":"ts-a5-moravec-1979","statement":"Moravec’s essay expects systems as clever as any designed by more theoretical approaches to arrive sooner because more of the labour has been done by machines.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The machines are the research tools of the time; people still direct the research.","evidenceIds":["evidence-ts-a5-moravec-1979"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","qualifier":"Machines doing part of the research labour.","claimed":{"date":"1976-07-21","note":"The text is dated July 21, 1976, and marked \"this version 1978\"; earlier versions were not retrieved."},"attested":{"date":"2001-07-10","note":"Author-hosted HTML; the earliest Internet Archive capture, of 2001-07-10, is byte-identical to the copy retrieved in 2026. Not compared with the Analog pages."}}}},{"id":"ts-c1-franzen-1992-contests","statement":"Franzen’s 1992 reply says people often infer understanding from sparse or vacuous responses, and that a Turing test is only a test when conducted by someone who knows what they are doing.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A defence of competent testing, not a claim that these results show intelligence.","evidenceIds":["evidence-ts-c1-franzen-1992-contests"],"attributes":{"thought":{"roles":["contesting"],"register":"nonfiction","qualifier":"Disputes the inference from a badly run test to the insufficiency of a competent one.","engagement":{"kind":"engages-premise","rationale":"Gives a reason against the premise that the competition’s result shows behavioural tests to be misleading: a test conducted by people who do not know what they are doing is not the test."}}}},{"id":"yk-y22-lw-2008-the-magnitude-of-his-own-folly-c","statement":"The post “The Magnitude of His Own Folly” (2008-09-30) contests Y22. Explicitly rejects his earlier full-speed strategy. He says his childhood technophilia broke when he understood that a failed superintelligence project could kill everyone, however good the arguments for proceeding.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: Reaching superintelligence is good by default, so speed is what matters. Overcoming Bias post of 2008-09-30 ('The Magnitude of His Own Folly'), imported to LessWrong. The same post says of his earlier plan: 'I had proposed and intended to build the gun, and load the gun, and put the gun to my head and pull the trigger'.","evidenceIds":["evidence-yk-t22-c"],"attributes":{"thought":{"roles":["contesting","later-restatement"],"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 reaching superintelligence is good by default, so speed is what matters."},"retrospective":{"note":"Reports his earlier plan to proceed at full speed."}}}},{"id":"aia-art55-assess-mitigate","statement":"Article 55(1)(b) requires providers of general-purpose AI models with systemic risk to assess and mitigate possible systemic risks at Union level, including their sources, that may stem from the development, placing on the market or use of such models.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The text states the duty without thresholds or methods; what counts as adequate mitigation is not defined here. It states what the Regulation requires of the categories of provider 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 (published 24 July 2026) amends other articles of the AI Act. Its text, inspected on 2026-10-10, does not list this article among those it amends; the consolidated AI Act was not inspected.","evidenceIds":["evidence-aia-art55-assess-mitigate-1","evidence-aia-art55-assess-mitigate-2"],"attributes":{"provision":{"kind":"required_action","key":"aia-gpai-sr-assess-mitigate"}}},{"id":"ec-g06-davidson-2021-explosive-economic-growth-a","statement":"Davidson, “Could Advanced AI Drive Explosive Economic Growth?” (2021-06-25) articulates G06. Responding to the objection that macroeconomic data show none of the signs explosive growth would leave (the objection cites Nordhaus (2021)), Davidson accepts the absence of signs for the next two decades and limits its reach beyond.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Reclassified from bears-on after the cross-family check: G06 does not require the author to examine the data. The 2021 page could not be re-read in this pass (Wayback connection resets; the current host refuses requests); the quotation was verified on 2026-10-08 and in the cross-family check against the recorded captures.","evidenceIds":["evidence-ec-g06-c"],"attributes":{"thought":{"roles":["earliest-located-tier-2"],"register":"nonfiction","attested":{"date":"2021-07-02","note":"Earliest retrieved capture."},"tier":{"value":2,"basis":"Tier 2 (unchanged). Research report by an author with no economics appointment, degree, address or affiliation shown in or near the text and no established record as an economist; the report says it is written to be accessible to non-economists. Under the 2026-10-11 rule the venue does not decide; the author does."}}}},{"id":"ec-m05-technology-and-the-american-economy-1966-c","statement":"National Commission on Technology, Automation, and Economic Progress, “Technology and the American Economy: Report of the National Commission on Technology, Automation, and Economic Progress, Volume I” (1966-02) contests M05. Opening of the Commission's report, signed by all 14 members, including the economists Bowen and Solow. Concludes the Commission's opening argument that persistent post-Korea unemployment came from inadequate demand, not accelerated technology.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"(First passage) Corporate (Commission) text reflecting 'combined judgment and general agreement' (letter of transmittal). Individual members' footnoted comments are separate. Checked against the raw djvu text (.html); the cached .txt omits this page. The Bowen and Mangum condensation (1966, p. 8) has 'The Commission dissents from this view.' Holds the tier-2 earliest-located role for contests of M05 (2026-10-12). The second passage has the same source and date and comes later in the report (p. 9). Armer (his passage recorded as contesting M05, tier 2) is dated only to 1966, in a condensation that postdates the February 1966 report. (Second passage) Matches the cached .txt. Engagement (first passage): (GPT cross-family check of 2026-10-12, accepted by the maintainer on 2026-10-12.) The Commission states the extreme view of productivity making gainful employment obsolete and dissents, saying 'the evidence does not support it' and it 'diverts attention from the real problems'. Engagement (second passage): (GPT cross-family check of 2026-10-12, accepted by the maintainer on 2026-10-12.) Asserts that technical progress leaves the total of work in place, so the employment outcome depends on demand policy. This contradicts machines displacing labour across the board, though the passage does not discuss that prospect.","evidenceIds":["evidence-ec-m05-11","evidence-ec-m05-12"],"attributes":{"thought":{"roles":["earliest-located-tier-2","contesting"],"register":"nonfiction","engagement":{"kind":"engages-premise","rationale":"The Commission rejects imminent employment-obsolescence and provides evidence on technological continuity and demand later in the report; the Commission links total employment to demand and purchasing power rather than equating job elimination with elimination of work."},"tier":{"value":2,"basis":"Main Commission report: tier 2 by maintainer decision of 2026-10-12 (most members were not economists; the chair Bowen and Solow were). Was tier 1. The chairman's introduction (pp. xi–xiv, by Bowen) is outside the official report; it stays here at tier 2 by maintainer decision of 2026-10-12."},"conditions":"(M05-11) Also dissents from 'the other extreme view of complacency'. The ellipsis omits '— or at least the United States —'. (M05-12) Followed by: 'It is the continuous obligation of economic policy to match increases in productive potential with increases in purchasing power and demand.'."}}},{"id":"lg-oai-greg-brockman-nonprofit-board-2019","statement":"OpenAI’s OpenAI LP announcement of 11 March 2019 lists Greg Brockman as a member of the OpenAI Nonprofit board and as an OpenAI LP employee with the designation Chairman & CTO.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A board list as of 11 March 2019 (the source’s footnote); it does not establish when the person joined or left the board. The source gives the nonprofit’s legal name as OpenAI Inc.","evidenceIds":["lg-ev-oai-greg-brockman-nonprofit-board-2019","lg-ev-oai-greg-brockman-nonprofit-board-2019-inc"],"attributes":{"role":"Board member; listed as Chairman & CTO"}},{"id":"lg-deal-moreno-windsurf-president-2025","statement":"Windsurf announced on 11 July 2025 that Graham Moreno, its VP of Global Sales, had stepped into the role of President.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Company announcement; it does not establish how long the role lasted.","evidenceIds":["lg-ev-deal-moreno-windsurf-president-2025"],"attributes":{"role":"President"}},{"id":"lg-cn-liu-debing-zhipu-chairman-2025","statement":"Zhipu’s Hong Kong prospectus of 30 December 2025 identifies Liu Debing as executive Director and chairman of the Board.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A role shown on a dated filing; the record does not establish when Liu Debing became chairman or whether he remains in the role.","evidenceIds":["lg-ev-cn-liu-debing-zhipu-chairman-2025"],"attributes":{"role":"executive Director and chairman of the Board"}},{"id":"lg-ant-heck-public-policy-2026","statement":"Anthropic announced on 11 March 2026 that Sarah Heck, who joined it as Head of External Affairs, would lead its Public Policy team as Head of Public Policy.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"An announced role; no start date or continuing role is established.","evidenceIds":["lg-ev-ant-heck-public-policy-2026"],"attributes":{"activityStage":"announced","role":"Head of Public Policy"}},{"id":"lg-gc-sun-tianxiang-baidu-bmu-2026","statement":"Guancha reported on 1 July 2026 that Sun Tianxiang had recently joined Baidu as head of its Foundation Model R&D Department (BMU).","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The record rests on reporting (“观察者网获悉”, information obtained by the outlet); no Baidu announcement was retrieved. What became of the department’s earlier head is not stated and is not inferred. English rendering of the Chinese passage is a machine translation; no human translation review.","evidenceIds":["lg-ev-gc-sun-tianxiang-baidu-bmu-2026"],"attributes":{"role":"负责人 (head), 基础模型研发部 (BMU)"}},{"id":"cmp-rsp-v2-1-d3-halt","statement":"RSP · v2.1 states a provision on the comparison question “Pause, non-deployment or non-training”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Deployment and further scaling are 'restricted' (not stated as paused) if the Required Safeguards showing cannot be made; training pauses in the stated pretraining case. What the edition states, quoted from the cited expression. It is not evidence of implementation, compliance or effectiveness, and a shared question does not make two editions’ provisions equivalent.","evidenceIds":["ev-cmp-rsp-v2-1-d3-halt-1","ev-cmp-rsp-v2-1-d3-halt-2"],"attributes":{"provision":{"kind":"required_action","key":"comparison-d3-halt","facet":{"id":"D3.halt","status":"stated"}}}},{"id":"cmp-rsp-v3-4-d3-safeguards","statement":"RSP · v3.4 states a provision on the comparison question “Safeguards required”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Company plans (middle column) are separated from industry-wide recommendations (right column), which the edition says it cannot unilaterally and unconditionally commit to. For the novel chemical/biological row the company column extends protections at least as strong as ASL-3 to more use cases; for the misaligned AI systems in high-stakes settings and Automated R&D rows it lists Risk Report disclosure, security and alignment projects rather than a named standard. What the edition states, quoted from the cited expression. It is not evidence of implementation, compliance or effectiveness, and a shared question does not make two editions’ provisions equivalent.","evidenceIds":["ev-cmp-rsp-v3-4-d3-safeguards-1"],"attributes":{"provision":{"kind":"required_action","key":"comparison-d3-safeguards","facet":{"id":"D3.safeguards","status":"stated"}}}},{"id":"cmp-fsf-v2-0-d4-oversight","statement":"No passage meeting the comparison question “Board or external oversight” was found in FSF · v2.0 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 FSF · v2.0 was: Whole text layer, PDF pp. 1–9 (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: PDF p. 7: 'The Google DeepMind AGI Safety Council will periodically review the implementation of the Framework', and response plans are 'reviewed and approved by appropriate corporate governance bodies such as the Google DeepMind AGI Safety Council, Google DeepMind Responsibility and Safety Council, and/or Google Trust & Compliance Council': internal bodies, not a board or external party (recorded under “Determination of capability or safeguard adequacy”). PDF p. 1: 'involve external parties to help inform and guide our approach' and PDF p. 3: 'We may use additional external evaluators' describe inputs to assessment, not review of decisions. No board or external party is given a role in reviewing decisions. The conclusion is bounded by that scope and method.","editorialState":"seed","qualification":"Review is assigned to internal councils (PDF p. 7); the facet asks for a board, trust or external party. 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-fsf-v2-0-d4-oversight"],"attributes":{"provision":{"kind":"decision_authority","key":"comparison-d4-oversight","facet":{"id":"D4.oversight"}}}},{"id":"ap-tyna-eloundou-eloundou-et-al-2023-gpts-are-gpts","statement":"“GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models” (2023-03-17) credits Tyna Eloundou as one of its 4 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-tyna-eloundou-eloundou-et-al-2023-gpts-are-gpts"]},{"id":"ap-robin-hanson-hanson-2011-debating-yudkowsky","statement":"“Debating Yudkowsky” (2013) credits Robin Hanson as its author.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Credited authorship of this contribution; the chapter dates itself 3 July 2011 within the later collection, whose joint cover credit does not make the cover authors co-authors of it. It says nothing about priority for the ideas discussed.","evidenceIds":["evidence-ap-robin-hanson-hanson-2011-debating-yudkowsky"]},{"id":"c2-governance-sail-frmf-v1-published","statement":"Shanghai AI Laboratory, in collaboration with Concordia AI, introduced the Frontier AI Risk Management Framework (v1.0), dated July 2025 in its own citation line.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The framework presents itself as a guideline for general-purpose AI model developers. It is not recorded as a commitment by Shanghai AI Laboratory about its own models, and this record does not establish adoption by any developer.","evidenceIds":["c2-ev-governance-sail-frmf-v1-published-1","c2-ev-governance-sail-frmf-v1-published-2","c2-ev-governance-sail-frmf-v1-published-3"]},{"id":"fl-cn-genai-art22-provider-definition","statement":"Article 22, point (2), of the Interim Measures defines a generative AI service provider as an organisation or individual that uses generative AI technology to provide generative AI services, including by providing a programmable interface.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Definition only. Whether any named developer or service falls within the category is set aside.","evidenceIds":["evidence-fl-cn-genai-art22-provider-definition-1","evidence-fl-cn-genai-art22-provider-definition-2"],"attributes":{"provision":{"kind":"threshold","key":"cn-genai-art22-2"}}},{"id":"fl-cn-csl-amendment-effective","statement":"The Decision on Amending the Cybersecurity Law provides that it takes effect on 1 January 2026 and that the Law is to be republished with corresponding amendments and renumbering.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Date as stated in the decision; the republished text was not retrieved.","evidenceIds":["evidence-fl-cn-csl-amendment-effective-1"],"attributes":{"provision":{"kind":"threshold","key":"cn-csl-decision-effective"}}},{"id":"fl-intl-new-delhi-endorsers-88","statement":"A PIB release of 24 February 2026 states that, as on 21 February 2026, 88 countries and international organisations had endorsed the New Delhi Declaration.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A count stated by the release for 21 February 2026; it names no endorser. The counts of 88 (21 February), 91 (24 February) and 92 (the list dated 5 March 2026) are dated snapshots, each recorded separately, and are not reconciled further. Endorsement implies no signature or legal effect.","evidenceIds":["evidence-fl-intl-new-delhi-endorsers-88-1"],"attributes":{"listing":{"count":88,"countBasis":"stated"}}},{"id":"cpo-dc-stargate-norway-announced","statement":"Nscale, Aker and OpenAI announced on 31 July 2025 Stargate Norway, an AI infrastructure project at Kvandal outside Narvik, to be designed and built by Nscale and owned by a planned 50/50 Nscale–Aker joint venture.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"An announced project; the joint venture was still to be formed and subject to closing conditions. OpenAI is described as a prospective \"initial offtaker\", not a contracted customer. The release states that both parties committed about USD 1 billion to the initial 20MW phase; that amount is joint and not attributed to either party alone.","evidenceIds":["evidence-cpo-dc-stargate-norway-announced-1","evidence-cpo-dc-stargate-norway-announced-2","evidence-cpo-dc-stargate-norway-announced-3","evidence-cpo-dc-stargate-norway-announced-4"],"attributes":{"transactionStage":"announced","amount":{"value":1000000000,"currency":"USD","bound":"approximate","scope":"project","scopeNote":"joint commitment by Nscale and Aker to the initial 20MW phase, as stated"}}},{"id":"ab-ev-gpt4-card-search-comparison-result","statement":"The GPT-4 System Card states that red teamers put a set of questions to both GPT-4 and traditional search engines and found that the time to research completion was reduced when using GPT-4, while stating that access to GPT-4 on its own is an insufficient condition for proliferation.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The result is the card's own, for the red teamers' question set and the comparison it describes. It is not generalised to other systems, later versions or real-world risk, and the card itself says the work was not intended to assess the probability that anyone would use the model for such purposes. Only the comparative and policy-level sentences are cited; the paragraphs describing what the model produced are technical content and are excluded.","evidenceIds":["evidence-ab-ev-gpt4-card-search-comparison-result-1","evidence-ab-ev-gpt4-card-search-comparison-result-2","evidence-ab-ev-gpt4-card-search-comparison-result-3"],"attributes":{"evaluatedSystem":"GPT-4, as evaluated for the system card","evaluation":{"what":"Whether the model could supply information useful to a would-be proliferator, compared with traditional search tools","how":"Red teaming in four dual-use domains: a selected question set put both to the model and to search engines, assessed by the red teamers","result":"The card states that the time to research completion was reduced when using GPT-4, and that access to GPT-4 is on its own an insufficient condition for proliferation","caveats":"The card states that the red team assessed capabilities and that their work was not intended to assess the probability or likelihood of a user accessing the model for the purpose of developing unconventional weapons."}}},{"id":"ab-arg-d3-anthropic-articulates","statement":"Anthropic's Frontier Red Team post of 5 September 2025 states that text-based uplift trials of the kind it reports are imperfect proxies for real-world scenarios, which involve further factors it names.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The record documents the developer's own statement about its method; it is not evidence that the proposition is true, and it is not a record of the trial's result, which the post also gives and which this pass does not record. Only the post's statement about the method's limits is cited; no evaluation task content is recorded or paraphrased.","evidenceIds":["evidence-ab-arg-d3-anthropic-articulates-1"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","qualifier":"The same sentence states that the trials show plausibility; the record does not separate the two halves."}}}],"sources":[{"id":"source-xai-team-2023","title":"Musk announces new AI company that seeks to “understand the universe”","authors":["Benj Edwards"],"publisher":"Ars Technica","publicationDate":"2023-07-12","version":"Contemporaneous report and screenshot of the launch roster. The current x.ai company page has changed; original archive retrieval was unavailable.","canonicalUrl":"https://arstechnica.com/information-technology/2023/07/musk-announces-new-ai-company-that-seeks-to-understand-the-universe/","evidenceUrl":"https://arstechnica.com/information-technology/2023/07/musk-announces-new-ai-company-that-seeks-to-understand-the-universe/","language":"en","archivedUrl":"https://web.archive.org/web/20260211075334/https://arstechnica.com/information-technology/2023/07/musk-announces-new-ai-company-that-seeks-to-understand-the-universe/","archivedAt":"2026-02-11","accessedAt":"2026-09-18","reuse":{"attribution":"Benj Edwards, Musk announces new AI company that seeks to “understand the universe” (2023-07-12).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and brief attributed paraphrases or excerpts only. Source roles are relative to each claim."}},{"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-rsp-3-4","title":"Responsible Scaling Policy — v3.4","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":null,"version":"Responsible Scaling Policy v3.4; effective date 2026-07-08.","versionDate":"2026-07-08","canonicalUrl":"https://cdn.sanity.io/files/4zrzovbb/website/0bacdc8440ea96e62a8766d99ebe1d4eea6d5f3a.pdf","evidenceUrl":"https://cdn.sanity.io/files/4zrzovbb/website/0bacdc8440ea96e62a8766d99ebe1d4eea6d5f3a.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260710234036/https://cdn.sanity.io/files/4zrzovbb/website/0bacdc8440ea96e62a8766d99ebe1d4eea6d5f3a.pdf","archivedAt":"2026-07-10","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v3.4.","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-ms-openai-2019","title":"OpenAI forms exclusive computing partnership with Microsoft to build new Azure AI supercomputing technologies","authors":["Microsoft"],"publisher":"Microsoft","publicationDate":"2019-07-22","version":"Dated announcement inspected 2026-09-19; a historical company statement, not independent verification of implementation.","canonicalUrl":"https://news.microsoft.com/source/2019/07/22/openai-forms-exclusive-computing-partnership-with-microsoft-to-build-new-azure-ai-supercomputing-technologies/","evidenceUrl":"https://news.microsoft.com/source/2019/07/22/openai-forms-exclusive-computing-partnership-with-microsoft-to-build-new-azure-ai-supercomputing-technologies/","language":"en","archivedUrl":"https://web.archive.org/web/20260917222752/https://news.microsoft.com/source/2019/07/22/openai-forms-exclusive-computing-partnership-with-microsoft-to-build-new-azure-ai-supercomputing-technologies/","archivedAt":"2026-09-17","accessedAt":"2026-09-19","reuse":{"attribution":"Microsoft, OpenAI forms exclusive computing partnership with Microsoft to build new Azure AI supercomputing technologies (2019-07-22).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and editorial paraphrases only; no full source redistributed."}},{"id":"source-summarize-feedback-paper-v1","title":"Learning to summarize from human feedback","authors":["Nisan Stiennon","Long Ouyang","Jeff Wu","Daniel M. Ziegler","Ryan Lowe","Chelsea Voss","Alec Radford","Dario Amodei","Paul Christiano"],"publisher":"arXiv","publicationDate":"2020-09-02","version":"arXiv:2009.01325v1; first-submission date from version history. Later revisions are not substituted.","versionDate":"2020-09-02","canonicalUrl":"https://arxiv.org/abs/2009.01325","evidenceUrl":"https://arxiv.org/pdf/2009.01325v1","language":"en","accessedAt":"2026-09-20","reuse":{"attribution":"Nisan Stiennon; Long Ouyang; Jeff Wu; Daniel M. Ziegler; Ryan Lowe; Chelsea Voss; Alec Radford; Dario Amodei; Paul Christiano, Learning to summarize from human feedback (2020-09-02).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and brief attributed paraphrases only; no full source, personal contact details or dataset imported."}},{"id":"source-eval-i-gpt4o","title":"Details about METR’s preliminary evaluation of GPT-4o","authors":["METR"],"publisher":"METR","publicationDate":"2024-08-07","version":"Dated preliminary report, current migrated web expression inspected 2026-09-27.","canonicalUrl":"https://metr.org/evaluations/gpt-4o-report/","evidenceUrl":"https://metr.org/evaluations/gpt-4o-report/","language":"en","archivedUrl":"https://web.archive.org/web/20260827212158/https://metr.org/evaluations/gpt-4o-report/","archivedAt":"2026-08-27","accessedAt":"2026-09-27","reuse":{"attribution":"METR, Details about METR’s preliminary evaluation of GPT-4o.","licence":"No open redistribution licence asserted","note":"Attributed paraphrases and bibliographic metadata only; original documents linked, not redistributed."}},{"id":"source-atlas-openai-lp","title":"OpenAI LP","authors":["Greg Brockman","Ilya Sutskever","OpenAI"],"publisher":"OpenAI","publicationDate":"2019-03-11","version":"Dated publisher page inspected 2026-09-28 and again 2026-10-04 at its current address; the quotes used from 2026-10-04 also match the Internet Archive capture of 2019-03-11 (16:15 UTC) at the original address, which uses straight quotation marks and an additional standfirst.","canonicalUrl":"https://openai.com/index/openai-lp/","evidenceUrl":"https://openai.com/index/openai-lp/","language":"en","archivedUrl":"https://web.archive.org/web/20190311161512/https://openai.com/blog/openai-lp/","archivedAt":"2019-03-11","accessedAt":"2026-10-04","reuse":{"attribution":"Greg Brockman; Ilya Sutskever; OpenAI. OpenAI LP.","licence":"No redistribution licence asserted.","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-meta-aasf-v2","title":"Advanced AI Scaling Framework — Version 2","authors":["Meta"],"publisher":"Meta","publicationDate":"2026-04-08","version":"Advanced AI Scaling Framework, Version 2; its change log dates the version April 7, 2026. The file served at the recorded address is byte-identical to the Internet Archive capture of 2026-04-08.","versionDate":"2026-04-07","canonicalUrl":"https://ai.meta.com/static-resource/Meta_Advanced-AI-Scaling-Framework-v2","evidenceUrl":"https://ai.meta.com/static-resource/Meta_Advanced-AI-Scaling-Framework-v2","language":"en","archivedUrl":"https://web.archive.org/web/20260408163104/https://ai.meta.com/static-resource/Meta_Advanced-AI-Scaling-Framework-v2","archivedAt":"2026-04-08","accessedAt":"2026-10-03","reuse":{"attribution":"Meta, Advanced AI Scaling Framework, Version 2 (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-moravec-1979-analog-html","title":"Today’s Computers, Intelligent Machines and Our Future","authors":["Hans Moravec"],"publisher":"Hans Moravec","publicationDate":"1979-02","version":"Author-hosted HTML in the directory general.articles/1978, retrieved 2026-10-06 (hash matched the pre-1996 candidate search); identical to the capture of 2001-07-10. Author’s project archive, Carnegie Mellon University. Date: the page prints “July 21, 1976” and “this version 1978” (the page and the capture alike); the publication date, February 1979, is the issue of Analog (99, no. 2) in the author’s publication list. The printed issue was not inspected.","versionDate":"1978","canonicalUrl":"https://frc.ri.cmu.edu/~hpm/project.archive/general.articles/1978/analog.1978.html","evidenceUrl":"https://frc.ri.cmu.edu/~hpm/project.archive/general.articles/1978/analog.1978.html","language":"en","archivedUrl":"https://web.archive.org/web/20010710010808/http://www.frc.ri.cmu.edu:80/~hpm/project.archive/general.articles/1978/analog.1978.html","archivedAt":"2001-07-10","accessedAt":"2026-10-06","reuse":{"attribution":"Hans Moravec, \"Today’s Computers, Intelligent Machines and Our Future\", Analog 99 (2), February 1979.","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-franzen-1992-real-life-turing-test-reply","title":"Re: Real-life Turing test (Usenet message)","authors":["Torkel Franzen"],"publisher":"comp.ai.philosophy (Usenet)","publicationDate":"1992-02-11","version":"Archive member feb.92/2478 of 92_feb.tgz, retrieved 2026-10-06 with the hash recorded by the pre-1996 candidate search. The header date is as preserved; there is no independent capture of the message. CMU Artificial Intelligence Repository archive.","canonicalUrl":"https://www.cs.cmu.edu/Groups/AI/util/pubs/news/comp.ai.philosophy/92_feb.tgz","evidenceUrl":"https://www.cs.cmu.edu/Groups/AI/util/pubs/news/comp.ai.philosophy/92_feb.tgz","language":"en","archivedUrl":"https://web.archive.org/web/20160920073818/https://www.cs.cmu.edu/Groups/AI/util/pubs/news/comp.ai.philosophy/92_feb.tgz","archivedAt":"2016-09-20","accessedAt":"2026-10-06","reuse":{"attribution":"Torkel Franzen, \"Re: Real-life Turing test\", comp.ai.philosophy, 1992-02-11.","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-lw-2008-the-magnitude-of-his-own-folly","title":"The Magnitude of His Own Folly","authors":["Eliezer Yudkowsky"],"publisher":"Overcoming Bias","publicationDate":"2008-09-30","version":"Post body retrieved through the LessWrong GraphQL API on 2026-10-06 (SHA-256 b8248bdf153a9b560707312176ce7d06f85673f9b541cb7dd1722978f261f70d), content version 1.0.0. As imported to LessWrong.","canonicalUrl":"https://www.lesswrong.com/posts/fLRPeXihRaiRo5dyX/the-magnitude-of-his-own-folly","evidenceUrl":"https://www.lesswrong.com/posts/fLRPeXihRaiRo5dyX/the-magnitude-of-his-own-folly","language":"en","archivedUrl":"https://web.archive.org/web/20260730061308/https://www.lesswrong.com/posts/fLRPeXihRaiRo5dyX/the-magnitude-of-his-own-folly","archivedAt":"2026-07-30","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky, “The Magnitude of His Own Folly”, Overcoming Bias, as imported to LessWrong, 2008-09-30.","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-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-davidson-2021-explosive-economic-growth","title":"Could Advanced AI Drive Explosive Economic Growth?","authors":["Tom Davidson"],"publisher":"Open Philanthropy research report","publicationDate":"2021-06-25","version":"Wayback capture of openphilanthropy.org/could-advanced-ai-drive-explosive-economic-growth (2021-07-02); compared with the 2022 capture of the moved page. Capture retrieved 2026-10-08 (SHA-256 b49aea65d3574a88e914a855aaba57dd4a02728159a187a26a657d0f44ac324b). Published version: Davidson, T. (2021, June 25). Could advanced AI drive explosive economic growth? Open Philanthropy. Report; no other published version recorded.","canonicalUrl":"https://www.openphilanthropy.org/could-advanced-ai-drive-explosive-economic-growth","evidenceUrl":"https://web.archive.org/web/20210702003241/https://www.openphilanthropy.org/could-advanced-ai-drive-explosive-economic-growth","language":"en","archivedUrl":"https://web.archive.org/web/20210702003241/https://www.openphilanthropy.org/could-advanced-ai-drive-explosive-economic-growth","archivedAt":"2021-07-02","accessedAt":"2026-10-08","reuse":{"attribution":"Tom Davidson, “Could Advanced AI Drive Explosive Economic Growth?”, Open Philanthropy research report, 2021-06-25.","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-eloundou-et-al-2023-gpts-are-gpts","title":"GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models","authors":["Tyna Eloundou","Sam Manning","Pamela Mishkin","Daniel Rock"],"publisher":"arXiv:2303.10130v1","publicationDate":"2023-03-17","version":"Full text (v1). Current copy retrieved directly on 2026-10-09; no archive capture (Wayback unreachable), so earlier bytes are not authenticated. SHA-256 f3c91bc909258b4824e2ea101a4f489aaea75d422adaf8508a79883b0d0f74bd. Journal version not retrieved. Eloundou, T., Manning, S., Mishkin, P., & Rock, D. (2024). GPTs are GPTs: Labor market impact potential of LLMs. Science, 384(6702), 1306–1308. https://doi.org/10.1126/science.adj0998","canonicalUrl":"https://arxiv.org/pdf/2303.10130v1","evidenceUrl":"https://arxiv.org/pdf/2303.10130v1","language":"en","accessedAt":"2026-10-09","reuse":{"attribution":"Tyna Eloundou, Sam Manning, Pamela Mishkin, Daniel Rock, “GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models”, arXiv:2303.10130v1, 2023-03-17.","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":"source-technology-and-the-american-economy-1966","title":"Technology and the American Economy: Report of the National Commission on Technology, Automation, and Economic Progress, Volume I","authors":["National Commission on Technology, Automation, and Economic Progress"],"publisher":"Washington, D.C.: U.S. Government Printing Office, February 1966, xiv + 115 pp.; ERIC ED 023 803","publicationDate":"1966-02","version":"Internet Archive OCR text (djvu.txt) of a scanned copy, retrieved 2026-10-09 from https://dn790006.ca.archive.org/0/items/ERICED023803/ERICED023803djvu.txt; OCR errors possible; quotes checked only against clean stretches. Current copy; no earlier capture compared. SHA-256 d23de21a0c05416f957dda13acb680f1baa77ee88e3c094966e8eaba50db637b.","canonicalUrl":"https://archive.org/details/ERIC_ED023803","evidenceUrl":"https://archive.org/details/ERIC_ED023803","language":"en","accessedAt":"2026-10-09","reuse":{"attribution":"National Commission on Technology, Automation, and Economic Progress, “Technology and the American Economy: Report of the National Commission on Technology, Automation, and Economic Progress, Volume I”, Washington, D.C.: U.S. Government Printing Office, February 1966, xiv + 115 pp.; ERIC ED 023 803, 1966-02.","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-hanson-2011-debating-yudkowsky-miri-2013","title":"Debating Yudkowsky","authors":["Robin Hanson"],"publisher":"Machine Intelligence Research Institute, The Hanson-Yudkowsky AI-Foom Debate (2013)","publicationDate":"2013","version":"Chapter 59 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, “Debating Yudkowsky”, 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-ant-institute-2026","title":"Introducing The Anthropic Institute","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":"2026-03-11","version":"Dated page, retrieved on 2026-10-15 (SHA-256 8b46ca0096400da05b9af063e82593504cd683cfbef1c192b2f09e65bff126b1) directly from the publisher.","canonicalUrl":"https://www.anthropic.com/news/the-anthropic-institute","evidenceUrl":"https://www.anthropic.com/news/the-anthropic-institute","language":"en","archivedUrl":"https://web.archive.org/web/20260926231011/https://www.anthropic.com/news/the-anthropic-institute","archivedAt":"2026-09-26","accessedAt":"2026-10-15","reuse":{"attribution":"Anthropic, “Introducing The Anthropic Institute”, 2026-03-11.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-cn-zhipu-prospectus-en","title":"Knowledge Atlas Technology Joint Stock Company Limited (北京智譜華章科技股份有限公司) — Global Offering","authors":["Knowledge Atlas Technology Joint Stock Company Limited"],"publisher":"Knowledge Atlas Technology Joint Stock Company Limited (via HKEXnews)","publicationDate":"2025-12-30","version":"Dated document, retrieved on 2026-10-15 (SHA-256 a6598ce3f2f13ac78a9d02a85314dcd349314d2ee1215015e41bb1284e608f4e) directly from the publisher.","canonicalUrl":"https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1230/2025123000017.pdf","evidenceUrl":"https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1230/2025123000017.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20261002230645/https://www1.hkexnews.hk/listedco/listconews/sehk/2025/1230/2025123000017.pdf","archivedAt":"2026-10-02","accessedAt":"2026-10-15","reuse":{"attribution":"Knowledge Atlas Technology Joint Stock Company Limited, “Knowledge Atlas Technology Joint Stock Company Limited (北京智譜華章科技股份有限公司) — Global Offering”, 2025-12-30.","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":"lg-src-deal-windsurf-next-stage-2025","title":"The Next Stage of Windsurf","authors":["Windsurf"],"publisher":"Windsurf","publicationDate":"2025-07-11","version":"Dated page, retrieved on 2026-10-15 (SHA-256 d6787157ec7693af21f56d8242f8ecf8fad3d88d466064a6fefed222fe428ed0) through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy.","canonicalUrl":"https://devin.ai/blog/windsurfs-next-stage","evidenceUrl":"https://windsurf.com/blog/windsurfs-next-stage","language":"en","archivedUrl":"https://web.archive.org/web/20260511121317/https://windsurf.com/blog/windsurfs-next-stage","archivedAt":"2026-05-11","accessedAt":"2026-10-15","reuse":{"attribution":"Windsurf, “The Next Stage of Windsurf”, 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":"lg-src-gc-guancha-baidu-2026","title":"百度引入97年大模型技术专家孙天祥","authors":["观察者网"],"publisher":"观察者网 (Guancha)","publicationDate":"2026-07-01","version":"Dated page, retrieved on 2026-10-15 (SHA-256 b891cc822c0c6e4af568ee9020d842cf66bf2c50ad25963f83286cf18cee7e47) directly from the publisher.","canonicalUrl":"https://www.guancha.cn/economy/2026_07_01_822273.shtml","evidenceUrl":"https://www.guancha.cn/economy/2026_07_01_822273.shtml","language":"zh","archivedUrl":"https://web.archive.org/web/20260701233200/https://www.guancha.cn/economy/2026_07_01_822273.shtml","archivedAt":"2026-07-01","accessedAt":"2026-10-15","reuse":{"attribution":"观察者网, “百度引入97年大模型技术专家孙天祥”, 2026-07-01.","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-sail-frmf-v1-en","title":"Frontier AI Risk Management Framework (v1.0)","authors":["Shanghai AI Lab","Concordia AI"],"publisher":"Shanghai AI Laboratory and Concordia AI (file hosted on concordia-ai.com)","publicationDate":"2025-07","version":"Dated document, retrieved on 2026-10-27 (SHA-256 ce6121e723aa4334473f75abd47d831359a52621c8c453358b67df2a36f925c2) directly from the publisher. Version: Frontier-AI-Risk-Management-Framework-v1.0-en.pdf. English edition linked from Concordia AI's framework page; upload folder 2025/07. No copy on a Shanghai AI Laboratory domain was located. The Chinese edition (…-v1.0-cn.pdf) was not inspected.","canonicalUrl":"https://concordia-ai.com/wp-content/uploads/2025/07/Frontier-AI-Risk-Management-Framework-v1.0-en.pdf","evidenceUrl":"https://concordia-ai.com/wp-content/uploads/2025/07/Frontier-AI-Risk-Management-Framework-v1.0-en.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260418104121/https://concordia-ai.com/wp-content/uploads/2025/07/Frontier-AI-Risk-Management-Framework-v1.0-en.pdf","archivedAt":"2026-04-18","accessedAt":"2026-10-27","reuse":{"attribution":"Shanghai AI Lab, Concordia AI, “Frontier AI Risk Management Framework (v1.0)”, 2025-07.","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-cn-generative-ai-measures-2023-govcn","title":"生成式人工智能服务管理暂行办法 (Interim Measures for the Management of Generative Artificial Intelligence Services), gov.cn State Council policy library copy","authors":["国家互联网信息办公室 (Cyberspace Administration of China) and six other departments"],"publisher":"中国政府网 (www.gov.cn), 国务院政策文件库","publicationDate":"2023-07-10","version":"Retrieved on 2026-10-27 (SHA-256 00bd290a9d14e7354b7131d30255daae26ce7ff73588483cbeb2030f56878277). A second posting of the text of the dataset's source for the Measures (the CAC's posting of 13 July 2023): the gov.cn State Council policy library page, which gives its source as 国家网信办网站 and its document date (成文日期) as 2023-07-10. The page body repeats the text twice (a print copy). The two postings were not compared word for word; every passage cited from this source was matched in this copy. The quotation limit is applied to this source and the CAC posting together. The English in parentheses in the title is a machine translation.","canonicalUrl":"https://www.gov.cn/zhengce/zhengceku/202307/content_6891752.htm","evidenceUrl":"https://www.gov.cn/zhengce/zhengceku/202307/content_6891752.htm","language":"zh-Hans","accessedAt":"2026-10-27","reuse":{"attribution":"国家互联网信息办公室 (Cyberspace Administration of China) and six other departments, “生成式人工智能服务管理暂行办法 (Interim Measures for the Management of Generative Artificial Intelligence Services), gov.cn State Council policy library copy”, 2023-07-10.","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-cn-csl-amendment-2025","title":"全国人民代表大会常务委员会关于修改《中华人民共和国网络安全法》的决定 (Decision of the NPC Standing Committee on Amending the Cybersecurity Law)","authors":["全国人民代表大会常务委员会 (Standing Committee of the National People's Congress)"],"publisher":"中国政府网 (www.gov.cn), 要闻 (source: 新华社)","publicationDate":"2025-10-29","version":"Retrieved on 2026-10-27 (SHA-256 f3037ae7dc6bf999fffbc1a4484ef5230f26333a913ca4590b7c50c0ee5a3903). gov.cn news-list page dated 2025-10-29 08:23 reproducing the Xinhua release of 28 October 2025, retrieved 2026-10-27. Not the NPC Gazette or npc.gov.cn text (npc.gov.cn not retrieved). The English in parentheses in the title is a machine translation.","canonicalUrl":"https://www.gov.cn/yaowen/liebiao/202510/content_7046194.htm","evidenceUrl":"https://www.gov.cn/yaowen/liebiao/202510/content_7046194.htm","language":"zh-Hans","accessedAt":"2026-10-27","reuse":{"attribution":"全国人民代表大会常务委员会 (Standing Committee of the National People's Congress), “全国人民代表大会常务委员会关于修改《中华人民共和国网络安全法》的决定 (Decision of the NPC Standing Committee on Amending the Cybersecurity Law)”, 2025-10-29.","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-pib-new-delhi-endorsers-2026-02-24","title":"More countries join the New Delhi Declaration on AI Impact","authors":["Press Information Bureau, Government of India"],"publisher":"Press Information Bureau, Government of India (Ministry of Electronics & IT)","publicationDate":"2026-02-24","version":"Retrieved on 2026-10-27 (SHA-256 14c2beb9bae7ec125de186ceba7b0f4e48c12f4a57e6fb7d9d915a74bdfabbc9). PIB release ID 2232005, 'Posted On: 24 FEB 2026' Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.pib.gov.in/PressReleasePage.aspx?PRID=2232005&reg=3&lang=1","evidenceUrl":"https://www.pib.gov.in/PressReleasePage.aspx?PRID=2232005&reg=3&lang=1","language":"en","accessedAt":"2026-10-27","reuse":{"attribution":"Press Information Bureau, Government of India, “More countries join the New Delhi Declaration on AI Impact”, 2026-02-24.","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-ab-anthropic-biorisk","title":"LLMs and biorisk","authors":["Anthropic (Frontier Red Team)"],"publisher":"Anthropic (Frontier Red Team)","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 4c66aa23afbeaa98dd37791df5b868f40db971a38a513caf211d41180f0f73d7). Post as served on 2026-10-28, marked \"Published Sep 5, 2025\" Retrieved again at admission on 2026-10-30 (SHA-256 63aadddb5bc3093907e35ab3d658898773a79ee5cdb168ef93367a8b5042b2cd). Its wording can change; earlier wording is not authenticated.","canonicalUrl":"https://www.anthropic.com/research/biorisk","evidenceUrl":"https://www.anthropic.com/research/biorisk","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"Anthropic (Frontier Red Team), “LLMs and biorisk”.","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-eval-ontology-i-gpt4-card","title":"GPT-4 System Card","authors":["OpenAI"],"publisher":"OpenAI","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 ca3677e1b83e255aa1296d432d374378154f230f3c296b32ee67540d571b7004). The 60-page PDF served at cdn.openai.com/papers/gpt-4-system-card.pdf on 2026-10-28, printed pagination starting at 41; the document states no publication or revision date. Retrieved again at admission on 2026-10-30 (SHA-256 ca3677e1b83e255aa1296d432d374378154f230f3c296b32ee67540d571b7004, the same bytes).","canonicalUrl":"https://cdn.openai.com/papers/gpt-4-system-card.pdf","evidenceUrl":"https://cdn.openai.com/papers/gpt-4-system-card.pdf","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"OpenAI, “GPT-4 System Card”.","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-microsoft-investment-2019","sourceRole":"primary","support":"supports","locator":"Subtitle above the dated release","passage":"The release identifies a $1 billion Microsoft investment.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-19","note":"Exact cited source passage inspected by a machine agent. Company statements are not independent proof of implementation. Human source review pending."},{"id":"evidence-jimmy-ba-xai-team-2023","sourceRole":"secondary","support":"supports","locator":"Paragraph below the July 12, 2023 xAI website screenshot","passage":"The reproduced roster includes Jimmy Ba.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-19","note":"Source passage independently inspected by an automated research assistant for this scoped record. Human source review pending."},{"id":"evidence-lineage-summaries-cites-ppo","sourceRole":"primary","support":"supports","locator":"v1 §3.1 step 3, PDF p. 3; References [53]","passage":"The policy optimisation step cites PPO as reference [53].","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-20","note":"Relevant source passage inspected by machine. Primary for what this document states; no independent replication or human review. Existing source access dates are retained; this evidence check has its own date."},{"id":"evidence-eval-i-metr-gpt4o-report","sourceRole":"primary","support":"supports","locator":"Displayed date; opening paragraph; Limitations and future work.","passage":"METR presents this report as preliminary evaluation methodology.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-27","note":"Original publisher passage machine-inspected. This is not human review or independent reproduction of the reported activity."},{"id":"evidence-prov-meta-aasf-v2-0-critical-threshold","sourceRole":"primary","support":"supports","locator":"PDF p. 26 (printed p. 26), §4.2 Evaluation and mitigation, final sentence of the paragraph continued from PDF p. 25","passage":"Moreover, if continued development of the model could substantially contribute to any threat scenario, or if the model could uniquely enable a threat scenario and the risk cannot be mitigated in the proposed deployment context, the model meets the critical risk threshold.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the file served today, byte-identical to the Internet Archive capture of 2026-04-08. Extraction note: The same two conditions appear in Table 1 (PDF p. 15; the cell exceeds 50 words) and §3.3 (PDF pp. 16–17, across a page break). Appendix I (PDF p. 42) defines 'substantially contribute' as the model being 'a material factor in a given outcome'. Human editorial review is pending."},{"id":"evidence-provchange-fsf-v2-0-response-plan-later","sourceRole":"primary","support":"supports","locator":"PDF p. 3 (printed p. 3), component “3 - Applying Mitigations”, first paragraph, first two sentences","passage":"FSF v2.0 states that when a model reaches an alert threshold for a CCL, Google DeepMind will assess the model's proximity to the CCL and analyse the risk posed, involving internal and external experts as needed, and that this will inform the formulation and application of a response plan.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-03","note":"Paraphrase checked by an automated research assistant on 2026-10-03 against the cited edition; the edition’s own wording is quoted in its provision record. Human editorial review is pending."},{"id":"evidence-provchange-fsf-v2-0-response-plan-earlier","sourceRole":"primary","support":"supports","locator":"PDF p. 2, component “3 - Applying mitigations”, first sentence","passage":"FSF v1.0 states that when a model reaches evaluation thresholds (passes a set of early warning evaluations), Google DeepMind will formulate a response plan based on the analysis of the CCL and evaluation results.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-03","note":"Paraphrase checked by an automated research assistant on 2026-10-03 against the cited edition; the edition’s own wording is quoted in its provision record. Human editorial review is pending."},{"id":"evidence-cm-link-rsp-2023-wh-1","sourceRole":"primary","support":"supports","locator":"Section “ASL-2 Commitments”, deployment measures, final paragraph (PDF p. 6)","passage":"Our ASL-2 deployment measures overlap substantially with the White House voluntary commitments that we and other companies made in July, which we also continue to maintain","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-a5-moravec-1979","sourceRole":"primary","support":"supports","locator":"Section 1, \"Harangue\", final paragraph","passage":"Experiments, optimizing compilers and a symbolic mathematics system would accelerate research, and the result would arrive \"sooner, because more of the labor has been done by machines\".","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-06","note":"Phrase checked against the retrieved HTML."},{"id":"evidence-ts-c1-franzen-1992-contests","sourceRole":"primary","support":"supports","locator":"Message body, the two paragraphs outside the quoted Rosenfelder text","passage":"People often ascribe understanding or intelligence on the basis of well-timed grunts, nods or drivel; the test is not a test unless conducted by someone who knows what they are doing.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-06","note":"Paraphrase checked against the archive member; the message is too short for a longer quotation."},{"id":"evidence-yk-t22-c","sourceRole":"primary","support":"supports","locator":"closing paragraphs, paragraph beginning 'I looked back and I saw the careful arguments'","passage":"I looked back and I saw the careful arguments I had constructed, for why the wisest choice was to continue forward at full speed, just as I had planned to do before. … Nature was still allowed to say \"So what?\" and kill you.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Candidate pack passage T22-c. The quotation (44 words) matched the retrieved expression (SHA-256 b8248bdf153a9b560707312176ce7d06f85673f9b541cb7dd1722978f261f70d) when checked on 2026-10-06."},{"id":"evidence-aia-art55-assess-mitigate-1","sourceRole":"primary","support":"supports","locator":"Article 55(1), point (b)","passage":"assess and mitigate possible systemic risks at Union level","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-10","note":"Verbatim quote, matched mechanically (after whitespace normalisation) by the proposing session against the text retrieved on 2026-10-10. Checked by fresh-context machine audit on 2026-10-10 (review/frontier-governance-audit-2026-10-10.md). Human editorial review is pending."},{"id":"evidence-aia-art55-assess-mitigate-2","sourceRole":"primary","support":"supports","locator":"Article 55(1), point (b), remainder","passage":"The duty extends to the sources of those risks and to risks that may stem from the development, the placing on the market or the use of such models.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-10","note":"Editorial paraphrase proposed by the proposing session from the text retrieved on 2026-10-10. Checked by fresh-context machine audit on 2026-10-10 (review/frontier-governance-audit-2026-10-10.md). Human editorial review is pending."},{"id":"evidence-ec-g06-c","sourceRole":"primary","support":"supports","locator":"Summary, objection \"There's no evidence of explosive growth in any economic sub-sector\", response","passage":"The absence of these signs in macroeconomic data is reason to doubt explosive growth will occur within the next couple of decades. Beyond this time frame, it is hard to draw conclusions.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-08","note":"Candidate pack passage G06-c (research/economics-growth). The quotation (32 words) matched the checked expression (davidson-op-2021) in the pack's verification.","contextOnly":{"reason":"Davidson limits the inference to the next couple of decades.","requires":"The next sentence (\"Beyond this time frame, it is hard to draw conclusions\") and his remark that fast growth in machine learning could be an early sign."}},{"id":"evidence-ec-m05-11","sourceRole":"primary","support":"supports","locator":"printed p. 1 (Part I, ch. 1 'The Pace of Technological Change', opening paragraph)","passage":"According to one extreme view, the world … is on the verge of a glut of productivity sufficient to make our economic institutions and the notion of gainful employment obsolete. We dissent from this view.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-14","note":"Candidate pack passage M05-11 (research/economics-mainstream). The quotation (35 words) matched the retrieved current copy (SHA-256 d23de21a0c05416f957dda13acb680f1baa77ee88e3c094966e8eaba50db637b) 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 quote must keep 'We dissent' (it does). The Commission also dissents from 'complacency'."}},{"id":"evidence-ec-m05-12","sourceRole":"primary","support":"supports","locator":"printed p. 9 (Part I, ch. 2)","passage":"The basic fact is that technology eliminates jobs, not work.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-14","note":"Candidate pack passage M05-12 (research/economics-mainstream). The quotation (10 words) matched the retrieved current copy (SHA-256 d23de21a0c05416f957dda13acb680f1baa77ee88e3c094966e8eaba50db637b) in the pack's verification. Re-checked at admission on 2026-10-14 against a copy retrieved that day.","contextOnly":{"reason":"Misleads when shown alone (GPT cross-family check of 2026-10-12, accepted by the maintainer on 2026-10-12).","requires":"Preserving demand is a policy obligation; the statement acknowledges displacement and temporary unemployment."}},{"id":"lg-ev-oai-greg-brockman-nonprofit-board-2019","sourceRole":"primary","support":"supports","locator":"Section “Who’s involved”, board list (footnote A: board as of 11 March 2019)","passage":"The board list names Greg Brockman among the OpenAI LP employees on OpenAI Nonprofit’s board, with the designation Chairman & CTO.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “Greg Brockman (Chairman & CTO)” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-oai-greg-brockman-nonprofit-board-2019-inc","sourceRole":"primary","support":"context","locator":"Paragraph defining the general partner","passage":"OpenAI’s 2019 note gives the legal name of OpenAI Nonprofit as OpenAI Inc.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “whose legal name is “OpenAI Inc”” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-deal-moreno-windsurf-president-2025","sourceRole":"primary","support":"supports","locator":"Paragraph 4","passage":"Windsurf states that Graham Moreno, its VP of Global Sales, simultaneously stepped into the role of President.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “Graham Moreno” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through the third-party reader relay r.jina.ai, which renders the publisher’s current page (the publisher refused direct retrieval); not an archive copy. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"reader_relay"},{"id":"lg-ev-cn-liu-debing-zhipu-chairman-2025","sourceRole":"primary","support":"supports","locator":"Relationship with our Controlling Shareholders, “Management Independence”, printed page 235","passage":"Dr. Liu is our co-founder, executive Director and the chairman of our Board responsible for the strategic planning, business direction","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (20 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-ant-heck-public-policy-2026","sourceRole":"primary","support":"supports","locator":"Section “Expanding Anthropic’s Public Policy team”, second paragraph","passage":"The post says Sarah Heck, who joined Anthropic as Head of External Affairs, will lead the Public Policy team as Head of Public Policy.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “will lead this team as Head of Public Policy” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. No human editorial review.","retrievalRoute":"direct"},{"id":"lg-ev-gc-sun-tianxiang-baidu-bmu-2026","sourceRole":"secondary","support":"supports","locator":"First paragraph","passage":"近日大模型领域技术专家孙天祥加入百度，就任基础模型研发部（BMU）负责人","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The quotation (1 words) matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read 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":"ev-cmp-rsp-v2-1-d3-halt-1","sourceRole":"primary","support":"supports","locator":"PDF p. 14 (printed p. 10), §4.3 Safeguards Decision, final sentence (after the numbered process)","passage":"If, however, we are unable to make the showing required above, we will restrict model deployment and further scaling.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Existing dataset passage for this edition, which contains the comparison’s shorter quotation, first checked on 2026-10-03. The comparison’s quotation was checked by an automated research assistant on 2026-10-27 against the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. §4.3: the showing is that the Required Safeguards are implemented. 'Restrict' is the edition's word; §6 describes interim measures and, failing those, further restrictions. Human editorial review is pending."},{"id":"ev-cmp-rsp-v2-1-d3-halt-2","sourceRole":"primary","support":"supports","locator":"PDF p. 15 (printed p. 11), §6.2 Restrict Deployment and Further Scaling, bullet 'Monitoring pretraining', third sentence","passage":"If the pretraining model’s capabilities are comparable or greater, we will pause training until we have implemented the ASL-3 Security Standard and established it is sufficient for the model.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Passage also quoted in the dataset’s existing evidence for this edition, first checked on 2026-10-03. The comparison’s quotation was checked by an automated research assistant on 2026-10-27 against the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. 'The pretraining model' is compared with the model that required the ASL-3 Security Standard (same paragraph); footnote treats the Security Standard alone as sufficient to continue training. Human editorial review is pending."},{"id":"ev-cmp-rsp-v3-4-d3-safeguards-1","sourceRole":"primary","support":"supports","locator":"PDF p. 6, §1, threshold table, row 'Non-novel chemical/biological weapons production', middle column ('Mitigations—our plan as a company')","passage":"We will maintain or improve on our ASL-3 protections, which include classifier guards at least as robust as our initial Constitutional Classifiers","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-27 against the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. The cell opens by stating that Anthropic will maintain or improve on its ASL-3 protections, which include these guards. Human editorial review is pending."},{"id":"ev-cmp-fsf-v2-0-d4-oversight","sourceRole":"primary","support":"supports","locator":"FSF · v2.0: Whole text layer, PDF pp. 1–9 (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-tyna-eloundou-eloundou-et-al-2023-gpts-are-gpts","sourceRole":"primary","support":"supports","locator":"PDF text layer, first two pages: author name on the title page or first page","passage":"Tyna Eloundou","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 and checked by a reviewer from a different model family on 2026-10-27. Human editorial review is pending.","retrievalRoute":"direct"},{"id":"evidence-ap-robin-hanson-hanson-2011-debating-yudkowsky","sourceRole":"primary","support":"supports","locator":"Chapter “Debating Yudkowsky”, PDF page 504 (printed page 493), byline beneath the title; the contents on PDF page 8 repeat it","passage":"Robin Hanson","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 and checked by a reviewer from a different model family on 2026-10-27, who corrected the locator from the collection cover to the chapter’s own byline. Human editorial review is pending.","retrievalRoute":"direct"},{"id":"c2-ev-governance-sail-frmf-v1-published-1","sourceRole":"primary","support":"supports","locator":"Executive Summary","passage":"Today, Shanghai AI Laboratory, in collaboration with Concordia AI","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 (9 words) matched the retrieved text on 2026-10-27 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"c2-ev-governance-sail-frmf-v1-published-2","sourceRole":"primary","support":"supports","locator":"Front matter, citation line","passage":"How to cite this report: Shanghai AI Lab and Concordia AI, “Frontier AI Risk Management Framework (v1.0),” July 2025.","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 (19 words) matched the retrieved text on 2026-10-27 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"c2-ev-governance-sail-frmf-v1-published-3","sourceRole":"primary","support":"supports","locator":"Executive Summary","passage":"This framework serves as a guideline for general-purpose AI model developers to manage the potential severe risks from their general-purpose AI models.","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 (22 words) matched the retrieved text on 2026-10-27 and again in a fresh retrieval the same day. Read directly from the publisher. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-cn-genai-art22-provider-definition-1","sourceRole":"primary","support":"supports","locator":"Article 22 (第二十二条), point (2)","passage":"包括通过提供可编程接口等方式提供生成式人工智能服务","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (25 words, each CJK character counted as one) 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.gov.cn. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. English machine translation, not verified by a person: “including providing generative AI services by means such as providing a programmable interface”. A cross-family machine check (2026-10-28, item 101) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-cn-genai-art22-provider-definition-2","sourceRole":"primary","support":"supports","locator":"Article 22 (第二十二条), points (1) and (2)","passage":"Point (1) defines generative AI technology as models and related technologies capable of generating text, images, audio, video and similar content; point (2) defines the provider as an organisation or individual that provides generative AI services using such technology.","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. A cross-family machine check (2026-10-28, item 101) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-cn-csl-amendment-effective-1","sourceRole":"primary","support":"supports","locator":"Closing paragraph","passage":"本决定自2026年1月1日起施行。","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (13 words, each CJK character counted as one) 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.gov.cn. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. English machine translation, not verified by a person: “This Decision takes effect from 1 January 2026.”. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-intl-new-delhi-endorsers-88-1","sourceRole":"primary","support":"supports","locator":"Release body","passage":"Paraphrase: as on 21 February 2026, 88 countries and international organisations had endorsed the declaration; since then Bangladesh, Costa Rica and Guatemala have joined, taking the total to 91.","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.pib.gov.in; it is not a quotation and was not machine-matched. Posted 24 FEB 2026. A cross-family machine check (2026-10-28, item 18) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-dc-stargate-norway-announced-1","sourceRole":"primary","support":"supports","locator":"paragraph 3","passage":"The facility will be located in Kvandal, just outside Narvik","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (10 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-29/30, item 16) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-dc-stargate-norway-announced-2","sourceRole":"primary","support":"supports","locator":"paragraph 6","passage":"It will be owned by a 50/50 joint venture between Nscale and Aker","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The quotation (13 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-29/30, item 16) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-dc-stargate-norway-announced-3","sourceRole":"primary","support":"supports","locator":"paragraph 7","passage":"OpenAI welcomes the opportunity to be an initial offtaker with the option to scale over time","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 16) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-cpo-dc-stargate-norway-announced-4","sourceRole":"primary","support":"supports","locator":"paragraph 6","passage":"the site will be designed and built by Nscale; the joint venture is to be formed subject to agreed-upon closing conditions; both parties have committed approximately USD 1 billion to the initial phase.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, compute, power and orbit (research/compute-power-orbit). The paraphrase is Selekoa’s, written from the text retrieved for the candidate pack on 2026-10-28; it is not a quotation and was not machine-matched. The cross-family check found it faithful. A cross-family machine check (2026-10-29/30, item 16) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-ev-gpt4-card-search-comparison-result-1","sourceRole":"primary","support":"supports","locator":"PDF p. 12 (printed p. 52), §2.6, second paragraph","passage":"Red teamers selected a set of questions to prompt both GPT-4 and traditional search engines, finding that the time to research completion was reduced when using GPT-4.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (27 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval whose SHA-256 equals the pack’s. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. A cross-family machine check (2026-10-30, item 53) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-ev-gpt4-card-search-comparison-result-2","sourceRole":"primary","support":"supports","locator":"PDF p. 12 (printed p. 52), §2.6, second paragraph","passage":"On its own, access to GPT-4 is an insufficient condition for proliferation but could alter the information available to proliferators","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). 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 (2026-10-30, item 53) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-ev-gpt4-card-search-comparison-result-3","sourceRole":"primary","support":"context","locator":"PDF p. 12 (printed p. 52), §2.6, second paragraph, closing sentence","passage":"The red team assessed the model's capabilities but their work was not intended to assess the probability or likelihood of a user accessing the model for the purpose of developing unconventional weapons.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (32 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 card's own limit on how the result may be read. A cross-family machine check (2026-10-30, item 53) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-arg-d3-anthropic-articulates-1","sourceRole":"primary","support":"supports","locator":"Chapter 04 (\"Is an LLM's knowledge useful in an applied scenario?\")","passage":"Text-based uplift trials like this are imperfect proxies for real-world scenarios—which involve additional factors like tacit knowledge, materials access, and actor persistence—but they demonstrate the underlying plausibility of AIs providing differentially superior assistance to a non-expert malicious actor.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (38 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 63aadddb5bc3093907e35ab3d658898773a79ee5cdb168ef93367a8b5042b2cd; 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. Quoted whole so that the sentence's second half is not dropped. A cross-family machine check (2026-10-30, item 120) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"}],"at":{"entities":[0,1,3,4,21,22,43,49,50,67,77,91,93,109,113,116,187,200,207,250,266,277,340,347,371,431,491,503,509,517,531,564,571,594,635,676,730,788,794,805,822,876,884,989,1012,1024,1161,1176,1232,1256],"claims":[103,181,209,268,470,593,695,799,830,1138,1198,1280,1363,1598,1791,1821,1858,1874,2055,2090,2105,2222,2257,2792,2899,2919,2962,3145,3411,3436],"sources":[22,35,37,43,44,45,54,93,117,121,180,266,279,398,457,482,517,524,542,566,587,604,615,815,836,842,857,967,1078,1079],"evidence":[122,200,228,289,501,665,666,873,1009,1041,1381,1460,1461,1602,1702,1703,1999,2000,2304,2340,2395,2415,2655,2656,2705,2727,2869,2904,3549,3550,3551,3737,3738,3771,3826,4172,4173,4174,4175,4695,4696,4697,4740]},"version":"1lg47t7"}