{"entities":[{"id":"openai","name":"OpenAI"},{"id":"anthropic","name":"Anthropic"},{"id":"amazon","name":"Amazon"},{"id":"eliezer-yudkowsky","name":"Eliezer Yudkowsky"},{"id":"rsp-2023","name":"Responsible Scaling Policy (2023)"},{"id":"iem-2013","name":"Intelligence Explosion Microeconomics"},{"id":"good-1965","name":"Speculations Concerning the First Ultraintelligent Machine"},{"id":"intelligence-explosion","name":"Intelligence explosion"},{"id":"google-research","name":"Google Research"},{"id":"rsp-v3-2","name":"Anthropic Responsible Scaling Policy — v3.2"},{"id":"fsf-v2-0","name":"Google DeepMind Frontier Safety Framework — v2.0"},{"id":"fsf-v3-0","name":"Google DeepMind Frontier Safety Framework — v3.0"},{"id":"fsf-v3-1","name":"Google DeepMind Frontier Safety Framework — v3.1"},{"id":"faif-2025","name":"xAI Frontier Artificial Intelligence Framework — December 2025 edition"},{"id":"jan-leike","name":"Jan Leike"},{"id":"miri-ml-agenda","name":"Alignment for Advanced Machine Learning Systems"},{"id":"meta","name":"Meta"},{"id":"llama-commercial-announcement","name":"Llama commercial-availability announcement (2023)"},{"id":"qwen-team","name":"Qwen Team / 通义千问"},{"id":"qwen3-policy","name":"Qwen3：思深，行速"},{"id":"meta-aasf-v2-0","name":"Meta Advanced AI Scaling Framework — Version 2"},{"id":"dsit","name":"Department for Science, Innovation and Technology"},{"id":"frontier-ai-safety-commitments-2024","name":"Frontier AI Safety Commitments, AI Seoul Summit 2024"},{"id":"rosenfelder-1992-real-life-turing-test","name":"Real-life Turing test"},{"id":"extropians-1996-08-30-clemmensen","name":"Re: Thinking about the future..."},{"id":"sl4-2002-03-08-yudkowsky","name":"The \"AI Box\" experiment"},{"id":"corrigibility-2015","name":"Corrigibility"},{"id":"christiano-2018-takeoff-speeds","name":"Takeoff speeds"},{"id":"agi-ruin-2022","name":"AGI Ruin: A List of Lethalities"},{"id":"recursive-self-improvement","name":"Recursive self-improvement"},{"id":"fast-or-hard-takeoff","name":"Fast or hard takeoff"},{"id":"treacherous-turn","name":"Treacherous turn"},{"id":"strategic-concealment","name":"Strategic concealment"},{"id":"behavioural-tests-as-evidence","name":"Behavioural tests as evidence of what a system is"},{"id":"warning-signs-in-time","name":"Clear warning signs will come before danger"},{"id":"diminishing-returns-to-intelligence","name":"Diminishing returns rule out a runaway"},{"id":"eu-ai-act-2024","name":"Regulation (EU) 2024/1689 (Artificial Intelligence Act), as published in the Official Journal"},{"id":"acemoglu-restrepo-2019-wrong-kind-of-ai","name":"The Wrong Kind of AI? Artificial Intelligence and the Future of Labor Demand"},{"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":"economics-engagement-with-superintelligence","name":"Economics’ engagement with superintelligence, AGI and the singularity"},{"id":"brynjolfsson-korinek-agrawal-2025-research-agenda-transformative-ai","name":"A Research Agenda for the Economics of Transformative AI"},{"id":"niki-parmar","name":"Niki Parmar"},{"id":"pamela-vagata","name":"Pamela Vagata"},{"id":"daron-acemoglu","name":"Daron Acemoglu"},{"id":"daniel-rock","name":"Daniel Rock"},{"id":"aiia-ai-safety-commitments-2024","name":"Artificial Intelligence Safety Commitments (人工智能安全承诺), December 2024"},{"id":"china-mobile","name":"China Mobile / 中国移动"},{"id":"uk-us-ai-safety-mou-2024","name":"UK–US Memorandum of Understanding on the science of AI safety (April 2024), as announced"},{"id":"uk-synthetic-nucleic-acids-screening-guidance-2024","name":"UK screening guidance on synthetic nucleic acids for users and providers"}],"claims":[{"id":"good-discusses-explosion-1965","statement":"Good’s paper discusses an intelligence explosion arising from machines designing better machines.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A historical formulation, not a claim of earliest use, inevitability, or transmission to a particular lab.","evidenceIds":["evidence-good-discusses-explosion-1965"]},{"id":"issued-rsp-3-2","statement":"Anthropic published Responsible Scaling Policy v3.2, identified by its effective date of 2026-04-29.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Publication means a public document expression exists. The date is labelled effective; publication is not evidence that its requirements were implemented or effective in practice.","evidenceIds":["evidence-issued-rsp-3-2"]},{"id":"lineage-ii-yudkowsky-agenda","statement":"Alignment for Advanced Machine Learning Systems credits Eliezer Yudkowsky as a co-author.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Authorship only. Work dated to 2016 by the publisher announcement; retrieved PDF revision is unknown.","evidenceIds":["evidence-lineage-ii-yudkowsky-agenda"]},{"id":"xm-llama-commercial","statement":"Meta announced Llama 2 availability for research and commercial use.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Company availability statement subject to the relevant licence terms; not a claim of unrestricted use, an external open-source certification or terms for later versions.","evidenceIds":["evidence-xm-llama-commercial"]},{"id":"cn-qwen3-license","statement":"Qwen Team’s Chinese Qwen3 announcement places its eight listed model weights under Apache 2.0.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"开源 is narrowed here to the specified weights and licence, not all data, all future Qwen models or an external open-source certification.","evidenceIds":["evidence-cn-qwen3-license"]},{"id":"provchange-fsf-v3-0-security-level-adjustment","statement":"v3.0's example for adjusting recommended security levels refers to models with capabilities not meaningfully different from other publicly available models that have weaker security applied, adding that the marginal benefit of higher security is then limited; v2.0's example referred to other widely available models that have not demonstrably caused or contributed to severe risks.","derivation":"inferred","resolution":"supported","inferenceRationale":"Premises: FSF · v3.0 states the provision as quoted in its provision record (PDF p. 9 (printed p. 9), §2.2 Misuse Critical Capability Levels, second paragraph, second sentence); FSF · v2.0 states it as quoted in its provision record (PDF p. 5 (printed p. 5), Misuse Critical Capability Levels, paragraph beginning “These recommended security levels”, second 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-v3-0-security-level-adjustment-later","evidence-provchange-fsf-v3-0-security-level-adjustment-earlier"],"attributes":{"changeBasis":"document_comparison","provision":{"kind":"exception","key":"fsf-security-level-adjustment"}}},{"id":"provchange-fsf-v3-1-deployment-mitigation-scope","statement":"v3.1's misuse deployment-mitigation criterion is stated to be required only for external deployment, not internal deployment or further development, and its ML R&D criterion for external and high-risk internal deployment; v3.0's misuse criterion was stated to be required only for external deployment, not further development, and its ML R&D criterion for external and large scale internal deployment.","derivation":"inferred","resolution":"supported","inferenceRationale":"Premises: FSF · v3.1 states the provision as quoted in its provision record (PDF p. 7 (printed p. 7), §1.3.5 Risk Acceptance Determination, misuse and ML R&D bullets, first sub-bullets, final sentences); FSF · v3.0 states it as quoted in its provision record (PDF p. 7 (printed p. 7), §1.6 Summary of Risk Acceptance Criteria, misuse and ML R&D bullets, first sub-bullets, final sentences). 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-v3-1-deployment-mitigation-scope-later","evidence-provchange-fsf-v3-1-deployment-mitigation-scope-earlier"],"attributes":{"changeBasis":"document_comparison","provision":{"kind":"exception","key":"fsf-deployment-mitigation-scope"}}},{"id":"cm-seoul-amazon","statement":"Amazon is listed among the sixteen organisations that, according to the GOV.UK text of 21 May 2024, the UK and Republic of Korea governments announced had agreed to the Frontier AI Safety Commitments.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Listing is the convener’s statement. It is not evidence that the organisation complied with, implemented or still maintains the commitments, and the legal effect of being listed is not recorded.","evidenceIds":["evidence-cm-seoul-amazon-1"],"attributes":{"commitment":{"stage":"listed"}}},{"id":"ts-c1-rosenfelder-1992","statement":"Rosenfelder’s 1992 message offers the restricted-topic Loebner competition in Boston, where, the message says, five of ten judges took a program for human, as food for thought for those who think the Turing test sufficient.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The five-of-ten figure is Rosenfelder’s report, not a verified result.","evidenceIds":["evidence-ts-c1-rosenfelder-1992"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction"}}},{"id":"ts-a4-clemmensen-1996-08-30-context","statement":"Clemmensen’s message of 1996-08-30 expects a superintelligence to design and build whatever automated tools it needs within a few days or weeks.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The period concerns the superintelligence’s use of people and tools, not the growth of its intelligence.","evidenceIds":["evidence-ts-a4-clemmensen-1996-08-30-context"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"The \"few days or weeks\" are for building tools once a superintelligence exists; no period is given for its rise from human level."}}},{"id":"ts-b2-yudkowsky-2002-aibox-context","statement":"Yudkowsky’s SL4 message of 2002-03-08 proposes a role-play in which the author plays a mind in a box and tries to persuade the other person to let it out.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A test of persuasion between humans; the transcript is kept secret.","evidenceIds":["evidence-ts-b2-yudkowsky-2002-aibox-context"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"A role-play of persuasion between two people, not a claim about AI systems."}}},{"id":"ts-b1-corrigibility","statement":"The Corrigibility paper describes an agent with incentives to deceive its programmers, maximising the intended utility only until it is powerful enough to avoid correction, and notes that Bostrom (2014) calls this a \"treacherous turn\". Recorded as articulating the proposition “Treacherous turn”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Primary for the paper’s own formal example and its citation; secondary for Bostrom’s formulation. The credit to Bostrom is the paper’s own.","evidenceIds":["evidence-ts-b1-corrigibility"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction"}}},{"id":"ts-a3-christiano-2018-takeoff","statement":"Christiano’s post says that before AI is great at self-improvement it will be mediocre at it, and that powerful AI can be used to develop better AI, leading to runaway growth.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The passage accepts recursive improvement; what it contests is the speed of the transition (A4).","evidenceIds":["evidence-ts-a3-christiano-2018-takeoff"],"attributes":{"thought":{"roles":["articulation"],"register":"nonfiction","qualifier":"Accepts recursive improvement while denying a discontinuity."}}},{"id":"yk-y08-agi-ruin-2022-c","statement":"The post “AGI Ruin: A List of Lethalities” (2022-06-05) contests Y08. Argues that some failure modes, such as deliberately appearing aligned to deceive operators, would first appear only at dangerous capability levels, so observing weaker systems cannot be relied on to surface them.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: Dangerous failure modes will show up first in weaker systems, where they can be observed and fixed.","evidenceIds":["evidence-yk-t08-d"],"attributes":{"thought":{"roles":["contesting"],"register":"nonfiction","engagement":{"kind":"engages-premise","rationale":"Under the message-level rule, the passage, read with the paragraph, message or dialogue turn it belongs to, gives a reason against the premise that dangerous failure modes will show up first in weaker systems, where they can be observed and fixed."}}}},{"id":"yk-y10-iem-2013-c","statement":"“Intelligence Explosion Microeconomics” (2013-09-13) contests Y10. Cites the hominid record as evidence against diminishing returns on cognitive improvement, and goes on to infer that researchers improving their own intelligence 'ought to go FOOM'.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Premise addressed: Returns to reinvesting cognitive gains in further cognitive improvement diminish. MIRI technical report 2013-1; cached PDF states 'Last modified September 13, 2013'. He adds in the same paragraph that this is hard to quantify.","evidenceIds":["evidence-yk-t10-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 returns to reinvesting cognitive gains in further cognitive improvement diminish."}}}},{"id":"ec-e05-brynjolfsson-korinek-agrawal-2025-research-agenda-transformative-ai-b","statement":"Brynjolfsson, Korinek and Agrawal, “A Research Agenda for the Economics of Transformative AI” (2025-09) bears on E05. Opening sentence of the research-agenda abstract ('As we approach Transformative Artificial Intelligence (TAI), …'). The agenda's list; 'safety risks' is one of nine challenges.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"(First passage) NBER w34256, September 2025; abstract page only. (Second passage) NBER w34256, September 2025; abstract page only. The abstract does not say what 'safety risks' covers; whether it includes existential risk was not checked against the full text.","evidenceIds":["evidence-ec-e05-29","evidence-ec-e05-30"],"attributes":{"thought":{"roles":["context"],"register":"nonfiction","missing":"(E05-29) Calls for an economics of transformative AI in general; the abstract does not name superintelligence, AGI, the singularity or existential risk. (E05-30) \"Safety risks\" in an abstract list; existential risk is not stated in the retrieved text.","tier":{"value":1,"basis":"NBER working paper (core venue); Brynjolfsson, Korinek and Agrawal are economists."}}}},{"id":"lg-oai-pamela-vagata-founding-member-2015","statement":"OpenAI’s launch announcement of 11 December 2015 names Pamela Vagata among the group’s founding members.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The predicate means participation in founding; the source wording is “founding members”, a group of research engineers and scientists, not a statement of legal founder status.","evidenceIds":["lg-ev-oai-pamela-vagata-founding-member-2015"],"attributes":{"role":"Founding member"}},{"id":"lg-goo-parmar-google-research-2017","statement":"Attention Is All You Need (arXiv v1, 12 June 2017) lists Niki Parmar with a Google Research affiliation.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"An author affiliation on a dated paper; it does not establish a title or tenure.","evidenceIds":["lg-ev-goo-parmar-google-research-2017"],"attributes":{"role":"author affiliation (Google Research)"}},{"id":"lg-oai-leike-departure-2024","statement":"Jan Leike posted on X on 17 May 2024 that the previous day had been his last day at OpenAI as head of alignment, superalignment lead and executive; he stated that he had been disagreeing with OpenAI leadership about the company’s core priorities for quite some time, until a breaking point was reached.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A first-person announcement. The stated reason is his own account in the same thread and is attributed, not adopted.","evidenceIds":["lg-ev-oai-leike-departure-2024-1","lg-ev-oai-leike-departure-2024-2"],"attributes":{"role":"Head of alignment, superalignment lead and executive","statedReason":{"by":"jan-leike","text":"Leike stated that he had been disagreeing with OpenAI leadership about the company’s core priorities for quite some time, until a breaking point was reached.","evidenceId":"lg-ev-oai-leike-departure-2024-2"}}},{"id":"cmp-meta-aasf-v2-0-d6-discretion","statement":"Meta AASF · v2 states a provision on the comparison question “Other exceptions or discretion”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Other stated room for judgement: cyber capabilities may be made available 'to qualified and trusted customers where appropriate' (PDF p. 30); chemical and biological capabilities may be offered to trusted researchers 'with reduced mitigations via structured access controls' (PDF p. 33, research area), and non-material modifications are presumed to inherit the underlying model's threshold (PDF p. 17). 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-meta-aasf-v2-0-d6-discretion-1"],"attributes":{"provision":{"kind":"exception","key":"comparison-d6-discretion","facet":{"id":"D6.discretion","status":"stated"}}}},{"id":"ap-daron-acemoglu-acemoglu-restrepo-2019-wrong-kind-of-ai","statement":"“The Wrong Kind of AI? Artificial Intelligence and the Future of Labor Demand” (2019-03) credits Daron Acemoglu as one of its 2 authors.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Credited co-authorship; it does not allocate parts of the text between authors or imply agreement with every passage.","evidenceIds":["evidence-ap-daron-acemoglu-acemoglu-restrepo-2019-wrong-kind-of-ai"]},{"id":"ap-daniel-rock-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 Daniel Rock 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-daniel-rock-eloundou-et-al-2023-gpts-are-gpts"]},{"id":"cmp-rsp-2023-d7-internal","statement":"RSP · 2023 states a provision on the comparison question “Internal reporting”.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"What the edition states, paraphrased from the cited expression with its page. It is not evidence of implementation, compliance or effectiveness, and a shared question does not make two editions’ provisions equivalent.","evidenceIds":["ev-cmp-rsp-2023-d7-internal-1"],"attributes":{"provision":{"kind":"reporting","key":"comparison-d7-internal","facet":{"id":"D7.internal","status":"stated"}}}},{"id":"cmp-faif-2025-d2-levels","statement":"FAIF · December 2025 edition states a provision on the comparison question “Named levels or thresholds”, in a weaker form than the question’s criterion.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Stated, partly: the passage has what the question’s criterion asks for, but in a weaker form than the criterion’s own words. No named or tiered capability levels. The edition states two numeric criteria for accepting the risk of deployment (this one and the MASK criterion under “Other named domains”) and says xAI plans to add further thresholds tied to other benchmarks. What the edition states, paraphrased from the cited expression with its page. It is not evidence of implementation, compliance or effectiveness, and a shared question does not make two editions’ provisions equivalent.","evidenceIds":["ev-cmp-faif-2025-d2-levels-1"],"attributes":{"provision":{"kind":"threshold","key":"comparison-d2-levels","facet":{"id":"D2.levels","status":"partly"}}}},{"id":"c2-governance-aiia-aisc-signatory-china-mobile","statement":"China Mobile is listed, as 中国移动通信集团有限公司, among the 17 companies on the initial signatory list (首批签署企业名单) of the AIIA AI Safety Commitments posted by CAICT on 24 December 2024.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"The list names the legal entity 中国移动通信集团有限公司. The record maps it to the group-level entity China Mobile, which carries that name as an alias; the mapping rests on the company name, and no claim is made about group structure or about which group company is bound. A listing is the publisher’s statement, dated by the post; it is not a signature date and not evidence that the company complied with or implemented the commitments. A later total of 22 signatories is reported, but no later list was retrieved.","evidenceIds":["c2-ev-governance-aiia-aisc-signatory-china-mobile-1","c2-ev-governance-aiia-aisc-signatory-china-mobile-2"],"attributes":{"commitment":{"stage":"listed"}}},{"id":"fl-uk-us-aisi-mou-signed","statement":"A GOV.UK press release of 2 April 2024 announces that the UK and US signed a Memorandum of Understanding, on 1 April 2024 as the release states, to work together on developing tests for the most advanced AI models, signed by Michelle Donelan and Gina Raimondo as Technology Secretary and US Commerce Secretary.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Announced, not inspected: the memorandum's text was not located. The release states intentions, including at least one joint testing exercise on a publicly accessible model; it does not show that such a test took place.","evidenceIds":["evidence-fl-uk-us-aisi-mou-signed-1","evidence-fl-uk-us-aisi-mou-signed-2"],"attributes":{"commitment":{"stage":"signed","scope":"UK–US cooperation between the AI Safety Institutes"}}},{"id":"ab-scr-uk-guidance-ai-in-background","statement":"The UK screening guidance states that increased access to synthetic nucleic acids, coupled with developments in converging tools and technologies such as AI and an increase in knowledge and expertise, may exacerbate the risks of engineering biology.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"Records what the guidance states as its background. It is not a finding that AI increases biological risk, and the guidance cites no evaluation or study for the statement in this passage. The provision kind is `risk_domain`, because the passage names the risk the guidance addresses rather than a duty; the schema has no `definition` kind.","evidenceIds":["evidence-ab-scr-uk-guidance-ai-in-background-1"],"attributes":{"provision":{"kind":"risk_domain","key":"uk-synthetic-nucleic-acids-guidance-ai-background"}}},{"id":"ab-gov-eu-ai-act-recital110-cbrn","statement":"Recital 110 of Regulation (EU) 2024/1689 states that international approaches have so far identified the need to pay attention to, among other risks of general-purpose AI models, chemical, biological, radiological and nuclear risks, such as the ways in which barriers to entry can be lowered, including for weapons development, design acquisition or use.","derivation":"explicit","resolution":"supported","editorialState":"seed","qualification":"A recital, which explains the Regulation rather than imposing a duty; the Regulation's articles are not recorded as naming this risk (recorded separately). It is not a finding that any model lowers any barrier, and it names no model or provider.","evidenceIds":["evidence-ab-gov-eu-ai-act-recital110-cbrn-1"],"attributes":{"provision":{"kind":"risk_domain","key":"eu-ai-act-recital110-cbrn-risks"}}}],"sources":[{"id":"source-openai-launch","title":"Introducing OpenAI","authors":["Greg Brockman","Ilya Sutskever","OpenAI"],"publisher":"OpenAI","publicationDate":"2015-12-11","version":"Dated page, inspected at retrieval","canonicalUrl":"https://openai.com/index/introducing-openai/","evidenceUrl":"https://openai.com/index/introducing-openai/","language":"en","archivedUrl":"https://web.archive.org/web/20260917013950/https://openai.com/index/introducing-openai/","archivedAt":"2026-09-17","accessedAt":"2026-09-17","reuse":{"attribution":"OpenAI, Introducing OpenAI (2015-12-11).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and minimal attributed excerpts only. Recheck terms before wider redistribution."}},{"id":"source-iem","title":"Intelligence Explosion Microeconomics","authors":["Eliezer Yudkowsky"],"publisher":"Machine Intelligence Research Institute","publicationDate":"2013","version":"Technical report 2013-1; last modified September 13, 2013","versionDate":"2013-09-13","canonicalUrl":"https://intelligence.org/files/IEM.pdf","evidenceUrl":"https://intelligence.org/files/IEM.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260917110159/https://intelligence.org/files/IEM.pdf","archivedAt":"2026-09-17","accessedAt":"2026-09-17","reuse":{"attribution":"Machine Intelligence Research Institute, Intelligence Explosion Microeconomics (2013).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and minimal attributed excerpts only. Recheck terms before wider redistribution."}},{"id":"source-good","title":"Speculations Concerning the First Ultraintelligent Machine","authors":["I. J. Good"],"publisher":"Academic Press","publicationDate":"1965","version":"Scanned reprint, Advances in Computers volume 6, pp. 31–88. Repository metadata identifies ©1965; the mirror filename says 1966.","canonicalUrl":"https://gwern.net/doc/ai/1966-good.pdf","evidenceUrl":"https://gwern.net/doc/ai/1966-good.pdf","language":"en","accessedAt":"2026-09-17","reuse":{"attribution":"Academic Press, Speculations Concerning the First Ultraintelligent Machine (1965).","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and minimal attributed excerpts only. Recheck terms before wider redistribution."}},{"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-3-2","title":"Responsible Scaling Policy — v3.2","authors":["Anthropic"],"publisher":"Anthropic","publicationDate":null,"version":"Responsible Scaling Policy v3.2; effective date 2026-04-29.","versionDate":"2026-04-29","canonicalUrl":"https://cdn.sanity.io/files/4zrzovbb/website/28c6241900d90410628a8a2003a5572faae4365a.pdf","evidenceUrl":"https://cdn.sanity.io/files/4zrzovbb/website/28c6241900d90410628a8a2003a5572faae4365a.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260505141510/https://cdn.sanity.io/files/4zrzovbb/website/28c6241900d90410628a8a2003a5572faae4365a.pdf","archivedAt":"2026-05-05","accessedAt":"2026-09-18","reuse":{"attribution":"Anthropic, Responsible Scaling Policy — v3.2.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-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-fsf-3-0","title":"Frontier Safety Framework — v3.0","authors":["Google DeepMind"],"publisher":"Google DeepMind","publicationDate":"2025-09-22","version":"Frontier Safety Framework v3.0; publication date 2025-09-22.","versionDate":"2025-09-22","canonicalUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3.pdf","evidenceUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20250922134959/https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3.pdf","archivedAt":"2025-09-22","accessedAt":"2026-09-18","reuse":{"attribution":"Google DeepMind, Frontier Safety Framework — v3.0.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-fsf-3-1","title":"Frontier Safety Framework — v3.1","authors":["Google DeepMind"],"publisher":"Google DeepMind","publicationDate":"2026-04-17","version":"Frontier Safety Framework v3.1; publication date 2026-04-17.","versionDate":"2026-04-17","canonicalUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf","evidenceUrl":"https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260423021420/https://storage.googleapis.com/deepmind-media/DeepMind.com/Blog/strengthening-our-frontier-safety-framework/frontier-safety-framework_3-1.pdf","archivedAt":"2026-04-23","accessedAt":"2026-09-18","reuse":{"attribution":"Google DeepMind, Frontier Safety Framework — v3.1.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-faif-2025","title":"Frontier Artificial Intelligence Framework — December 2025 edition","authors":["xAI"],"publisher":"xAI","publicationDate":null,"version":"Frontier Artificial Intelligence Framework December 2025 edition; revision date 2025-12-30.","versionDate":"2025-12-30","canonicalUrl":"https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf","evidenceUrl":"https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260102171656/https://data.x.ai/2025-12-31-xai-frontier-artificial-intelligence-framework.pdf","archivedAt":"2026-01-02","accessedAt":"2026-09-18","reuse":{"attribution":"xAI, Frontier Artificial Intelligence Framework — December 2025 edition.","licence":"No open redistribution licence asserted","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-attention-all-you-need-v1","title":"Attention Is All You Need","authors":["Ashish Vaswani","Noam Shazeer","Niki Parmar","Jakob Uszkoreit","Llion Jones","Aidan N. Gomez","Lukasz Kaiser","Illia Polosukhin"],"publisher":"arXiv","publicationDate":"2017-06-12","version":"arXiv:1706.03762v1; first-submission date from version history. Later revisions are not substituted.","versionDate":"2017-06-12","canonicalUrl":"https://arxiv.org/abs/1706.03762","evidenceUrl":"https://arxiv.org/pdf/1706.03762v1","language":"en","accessedAt":"2026-09-20","reuse":{"attribution":"Ashish Vaswani; Noam Shazeer; Niki Parmar; Jakob Uszkoreit; Llion Jones; Aidan N. Gomez; Lukasz Kaiser; Illia Polosukhin, Attention Is All You Need (2017-06-12).","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-miri-ml-agenda","title":"Alignment for Advanced Machine Learning Systems","authors":["Jessica Taylor","Eliezer Yudkowsky","Patrick LaVictoire","Andrew Critch"],"publisher":"Machine Intelligence Research Institute","publicationDate":"2016","version":"Publisher-hosted 25-page PDF inspected 2026-09-25. Work year corroborated by the 2016 announcement; PDF revision date unknown.","canonicalUrl":"https://intelligence.org/files/AlignmentMachineLearning.pdf","evidenceUrl":"https://intelligence.org/files/AlignmentMachineLearning.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20260922133238/https://intelligence.org/files/AlignmentMachineLearning.pdf","archivedAt":"2026-09-22","accessedAt":"2026-09-25","reuse":{"attribution":"Jessica Taylor; Eliezer Yudkowsky; Patrick LaVictoire; Andrew Critch, Alignment for Advanced Machine Learning Systems.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata and brief attributed paraphrases only; source publications and personal contact details are not redistributed."}},{"id":"source-xm-llama-commercial","title":"Meta and Microsoft Introduce the Next Generation of Llama","authors":["Meta"],"publisher":"Meta","publicationDate":"2023-07-18","version":"Dated publisher webpage as retrieved; not an authenticated historical capture.","canonicalUrl":"https://about.fb.com/news/2023/07/llama-2/","evidenceUrl":"https://about.fb.com/news/2023/07/llama-2/","language":"en","archivedUrl":"https://web.archive.org/web/20260921061736/https://about.fb.com/news/2023/07/llama-2/","archivedAt":"2026-09-21","accessedAt":"2026-09-29","reuse":{"attribution":"Meta and Microsoft Introduce the Next Generation of Llama — Meta","licence":"No redistribution licence asserted.","note":"Source-linked metadata and brief evidence paraphrases; original text remains with its publisher."}},{"id":"source-cn-qwen3","title":"Qwen3：思深，行速","authors":["Qwen Team"],"publisher":"Qwen Team","publicationDate":"2025-04-29","version":"Source language: Simplified Chinese (zh-Hans). Dated original Chinese team announcement; legacy site now carries a redirect notice. Body inspected at retrieval.","canonicalUrl":"https://qwenlm.github.io/zh/blog/qwen3/","evidenceUrl":"https://qwenlm.github.io/zh/blog/qwen3/","language":"zh-Hans","archivedUrl":"https://web.archive.org/web/20260513135107/https://qwenlm.github.io/zh/blog/qwen3/","archivedAt":"2026-05-13","accessedAt":"2026-09-29","reuse":{"attribution":"Qwen Team, Qwen3：思深，行速 (2025-04-29).","licence":"No redistribution licence asserted for this source expression","note":"Bibliographic metadata and brief attributed paraphrases only; model licensing is distinct from source-document reuse."}},{"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-fasc-2024","title":"Frontier AI Safety Commitments, AI Seoul Summit 2024","authors":["Department for Science, Innovation and Technology"],"publisher":"Department for Science, Innovation and Technology","publicationDate":"2024-05-21","version":"GOV.UK HTML publication, first published 21 May 2024 and updated 7 February 2025 to add organisations to the list; inspected live 2026-10-04. The Internet Archive capture of 2024-05-21 (11:05 UTC) has the sixteen-name list only, and the commitment text quoted is identical in both. Publisher as printed on the inspected expression: Department for Science, Innovation and Technology (GOV.UK).","versionDate":"2025-02-07","canonicalUrl":"https://www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024/frontier-ai-safety-commitments-ai-seoul-summit-2024","evidenceUrl":"https://www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024/frontier-ai-safety-commitments-ai-seoul-summit-2024","language":"en","archivedUrl":"https://web.archive.org/web/20240521110530/https://www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024/frontier-ai-safety-commitments-ai-seoul-summit-2024","archivedAt":"2024-05-21","accessedAt":"2026-10-04","reuse":{"attribution":"Department for Science, Innovation and Technology, Frontier AI Safety Commitments, AI Seoul Summit 2024 (GOV.UK).","licence":"Open Government Licence v3.0 (GOV.UK content); quoted briefly under the editorial policy","note":"Selekoa practice: metadata, attributed paraphrases and short attributed quotations under the editorial policy (at most 50 words each; at most 400 quoted words per document across the dataset, and about 10% of a document under about 2,000 words). Documents are linked externally, not redistributed. Publisher terms that forbid quotation override this practice."}},{"id":"source-rosenfelder-1992-real-life-turing-test","title":"Real-life Turing test (Usenet message)","authors":["Mark Rosenfelder"],"publisher":"comp.ai.philosophy (Usenet)","publicationDate":"1992-02-10","version":"Archive member feb.92/2462 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","accessedAt":"2026-10-06","reuse":{"attribution":"Mark Rosenfelder, \"Real-life Turing test\", comp.ai.philosophy, 1992-02-10.","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-extropians-1996-08-30-clemmensen","title":"Re: Thinking about the future... (Extropians mailing list)","authors":["Dan Clemmensen"],"publisher":"Extropians mailing list","publicationDate":"1996-08-30","version":"Archive HTML as served, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search; quoted lines excluded when reading the author’s text. Archive at extropians.weidai.com.","canonicalUrl":"http://extropians.weidai.com/extropians.96/0613.html","evidenceUrl":"http://extropians.weidai.com/extropians.96/0613.html","language":"en","archivedUrl":"https://web.archive.org/web/20150427133328/http://extropians.weidai.com/extropians.96/0613.html","archivedAt":"2015-04-27","accessedAt":"2026-10-06","reuse":{"attribution":"Dan Clemmensen, \"Re: Thinking about the future...\", Extropians mailing list, 1996-08-30.","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-sl4-2002-03-08-yudkowsky","title":"The \"AI Box\" experiment (SL4 mailing list)","authors":["Eliezer S. Yudkowsky"],"publisher":"SL4 mailing list","publicationDate":"2002-03-08","version":"Archive HTML as served, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search. The served archive shows later URL rewriting elsewhere, so it does not authenticate the exact historical bytes. Archive at sl4.org.","canonicalUrl":"http://sl4.org/archive/0203/3132.html","evidenceUrl":"http://sl4.org/archive/0203/3132.html","language":"en","archivedUrl":"https://web.archive.org/web/20070505151319/http://sl4.org:80/archive/0203/3132.html","archivedAt":"2007-05-05","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer S. Yudkowsky, \"The \"AI Box\" experiment\", SL4 mailing list, 2002-03-08.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-corrigibility-2015-pdf","title":"Corrigibility","authors":["Nate Soares","Benja Fallenstein","Eliezer Yudkowsky","Stuart Armstrong"],"publisher":"AAAI Workshops","publicationDate":"2015-01-25","version":"MIRI-hosted PDF of the workshop paper, retrieved 2026-10-06 with the hash recorded by the 1996–2022 candidate search.","canonicalUrl":"https://intelligence.org/files/Corrigibility.pdf","evidenceUrl":"https://intelligence.org/files/Corrigibility.pdf","language":"en","archivedUrl":"https://web.archive.org/web/20150322195302/https://intelligence.org/files/Corrigibility.pdf","archivedAt":"2015-03-22","accessedAt":"2026-10-06","reuse":{"attribution":"Soares, Fallenstein, Yudkowsky and Armstrong, \"Corrigibility\", AAAI Workshops (2015).","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-christiano-2018-takeoff-speeds","title":"Takeoff speeds","authors":["Paul Christiano"],"publisher":"The sideways view (Paul Christiano’s blog)","publicationDate":"2018-02-24","version":"Current page, retrieved 2026-10-06; its bytes differ from the 1996–2022 candidate search’s retrieval while the quoted text matches. A machine check compared the capture of 2018-02-25, which was retrieved again on 2026-10-06.","canonicalUrl":"https://sideways-view.com/2018/02/24/takeoff-speeds/","evidenceUrl":"https://sideways-view.com/2018/02/24/takeoff-speeds/","language":"en","archivedUrl":"https://web.archive.org/web/20180225225416/https://sideways-view.com/2018/02/24/takeoff-speeds/","archivedAt":"2018-02-25","accessedAt":"2026-10-06","reuse":{"attribution":"Paul Christiano, \"Takeoff speeds\", The sideways view, 24 February 2018.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-agi-ruin-2022","title":"AGI Ruin: A List of Lethalities","authors":["Eliezer Yudkowsky"],"publisher":"LessWrong and Alignment Forum","publicationDate":"2022-06-05","version":"Text checked in the post’s first stored version through the LessWrong GraphQL API on 2026-10-06.","canonicalUrl":"https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","evidenceUrl":"https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","language":"en","archivedUrl":"https://web.archive.org/web/20220605225854/https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a-list-of-lethalities","archivedAt":"2022-06-05","accessedAt":"2026-10-06","reuse":{"attribution":"Eliezer Yudkowsky, \"AGI Ruin: A List of Lethalities\", LessWrong and Alignment Forum, 2022-06-05.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-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-acemoglu-restrepo-2019-wrong-kind-of-ai","title":"The Wrong Kind of AI? Artificial Intelligence and the Future of Labor Demand","authors":["Daron Acemoglu","Pascual Restrepo"],"publisher":"NBER Working Paper 25682","publicationDate":"2019-03","version":"Full text. Current copy retrieved directly on 2026-10-09; no archive capture (Wayback unreachable), so earlier bytes are not authenticated. SHA-256 54ad101804ab7d30ba6d136f413b1c871ffec0a273dbec290c009acd7ad32100.","canonicalUrl":"https://www.nber.org/system/files/working_papers/w25682/w25682.pdf","evidenceUrl":"https://www.nber.org/system/files/working_papers/w25682/w25682.pdf","language":"en","accessedAt":"2026-10-09","reuse":{"attribution":"Daron Acemoglu, Pascual Restrepo, “The Wrong Kind of AI? Artificial Intelligence and the Future of Labor Demand”, NBER Working Paper 25682, 2019-03.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"source-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-brynjolfsson-korinek-agrawal-2025-research-agenda-transformative-ai","title":"A Research Agenda for the Economics of Transformative AI","authors":["Erik Brynjolfsson","Anton Korinek","Ajay Agrawal"],"publisher":"NBER Working Paper 34256","publicationDate":"2025-09","version":"NBER abstract page. Retrieved 2026-10-11. SHA-256 8fb8e191f7c3d57f9136f42e237a3b753aa23c2c4e7c09bd04df5b2568381666. Version a reader would cite, or full text, not retrieved. Brynjolfsson, E., Korinek, A., & Agrawal, A. (2025). A research agenda for the economics of transformative AI (NBER Working Paper No. 34256). https://doi.org/10.3386/w34256","canonicalUrl":"https://www.nber.org/papers/w34256","evidenceUrl":"https://www.nber.org/papers/w34256","language":"en","accessedAt":"2026-10-11","reuse":{"attribution":"Erik Brynjolfsson, Anton Korinek, Ajay Agrawal, “A Research Agenda for the Economics of Transformative AI”, NBER Working Paper 34256, 2025-09.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-oai-leike-priorities-x-2024","title":"Post of 17 May 2024 (second post of the thread on leaving OpenAI)","authors":["Jan Leike"],"publisher":"X","publicationDate":"2024-05-17","version":"Dated post, retrieved on 2026-10-15 (SHA-256 e3b87c36aa0125f9fafec84868e5ef8a4fd80eae9cc4d7ed65242612421175cc) through X’s own syndication endpoint for the post’s text, which truncates long notes. Read through X’s syndication endpoint (text only). Publication date is the post’s UTC timestamp.","canonicalUrl":"https://x.com/janleike/status/1791498178346549382","evidenceUrl":"https://x.com/janleike/status/1791498178346549382","language":"en","archivedUrl":"https://web.archive.org/web/20261012080506/https://x.com/janleike/status/1791498178346549382","archivedAt":"2026-10-12","accessedAt":"2026-10-15","reuse":{"attribution":"Jan Leike, “Post of 17 May 2024 (second post of the thread on leaving OpenAI)”, 2024-05-17.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"lg-src-oai-leike-x-2024","title":"Post of 17 May 2024 on his last day at OpenAI (first post of a thread)","authors":["Jan Leike"],"publisher":"X","publicationDate":"2024-05-17","version":"Dated post, retrieved on 2026-10-15 (SHA-256 5d96c342e2bab3675eb9b97853b35663a21e7daa3a90bd8dc61b7a097b4a81f8) through X’s own syndication endpoint for the post’s text, which truncates long notes. Read through X’s syndication endpoint (text only). Publication date is the post’s UTC timestamp.","canonicalUrl":"https://x.com/janleike/status/1791498174659715494","evidenceUrl":"https://x.com/janleike/status/1791498174659715494","language":"en","archivedUrl":"https://web.archive.org/web/20260919031016/https://x.com/janleike/status/1791498174659715494","archivedAt":"2026-09-19","accessedAt":"2026-10-15","reuse":{"attribution":"Jan Leike, “Post of 17 May 2024 on his last day at OpenAI (first post of a thread)”, 2024-05-17.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}},{"id":"c2-src-caict-wx-aisc-2024","title":"守护AI安全，共建行业自律典范——首批17家企业签署《人工智能安全承诺》","authors":["中国信通院CAICT"],"publisher":"中国信息通信研究院 (CAICT), WeChat official account 中国信通院CAICT","publicationDate":"2024-12-24","version":"Dated post, retrieved on 2026-10-27 (SHA-256 4a3b099c801423b12830598c841921a74d01d4d9d3f69379669aaf17e6c38d78) directly from the publisher. The commitment text (Chinese and English) and the initial signatory list are a single embedded image (1000 × 9898 pixels, SHA-256 7e9145350d1b882d22ab3f7fc4d4c0226f06194f94264d92291ee065fbf4aa27), not in the page’s text layer; quotations from it are transcriptions. The cross-family check found the same image in an Internet Archive capture of 4 March 2026 with the same SHA-256. Its wording can change; earlier wording is not authenticated. The cross-family check (2026-10-27) located the Internet Archive capture linked here; a capture shows only what the archive retrieved that day. English title (machine translation): “Safeguarding AI safety, building a model of industry self-discipline: initial 17 companies sign the AI Safety Commitments”.","canonicalUrl":"https://mp.weixin.qq.com/s/s-XFKQCWhu0uye4opgb3Ng","evidenceUrl":"https://mp.weixin.qq.com/s/s-XFKQCWhu0uye4opgb3Ng","language":"zh-Hans","archivedUrl":"https://web.archive.org/web/20260724145912/https://mp.weixin.qq.com/s/s-XFKQCWhu0uye4opgb3Ng","archivedAt":"2026-07-24","accessedAt":"2026-10-27","reuse":{"attribution":"中国信通院CAICT, “守护AI安全，共建行业自律典范——首批17家企业签署《人工智能安全承诺》”, 2024-12-24.","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":"c2-src-itcloudbd-aisc-2024","title":"守护AI安全，共建行业自律典范 首批17家企业签署《人工智能安全承诺》","authors":["科技云报到"],"publisher":"科技云报到 (itcloudbd.com)","publicationDate":"2024-12-27","version":"Dated page, retrieved on 2026-10-27 (SHA-256 f4d6790217cccacfc6fea787516f4538abb1d4c1bf03b65d3f5123407c0d72f4) directly from the publisher. Tech-media article dated 2024-12-27, read on 2026-10-27. It reproduces the CAICT post's wording and gives the commitments and signatory names as text; its commitment-two wording (红队测试) matches the image in the CAICT post. Its wording can change; earlier wording is not authenticated. The cross-family check (2026-10-27) located the Internet Archive capture linked here; a capture shows only what the archive retrieved that day. English title (machine translation): “First 17 companies sign the AI Safety Commitments”. A reporting outlet, cited beside the primary text.","canonicalUrl":"https://itcloudbd.com/rengongzhineng/13352.html","evidenceUrl":"https://itcloudbd.com/rengongzhineng/13352.html","language":"zh-Hans","archivedUrl":"https://web.archive.org/web/20260712204752/https://itcloudbd.com/rengongzhineng/13352.html","archivedAt":"2026-07-12","accessedAt":"2026-10-27","reuse":{"attribution":"科技云报到, “守护AI安全，共建行业自律典范 首批17家企业签署《人工智能安全承诺》”, 2024-12-27.","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-uk-us-ai-safety-mou-release-2024","title":"UK & United States announce partnership on science of AI safety","authors":["Department for Science, Innovation and Technology"],"publisher":"Department for Science, Innovation and Technology","publicationDate":"2024-04-02","version":"Retrieved on 2026-10-27 (SHA-256 46da83fc5bb30d6c542e9e0dbda534eabc797046fa53d43c222a2d3942fc3725). GOV.UK press release, 'Published: 2 April 2024'; the memorandum's own text was not located or inspected Its wording can change; earlier wording is not authenticated. Publisher as printed on the inspected expression: GOV.UK (Department for Science, Innovation and Technology, AI Safety Institute and the Secretary of State).","canonicalUrl":"https://www.gov.uk/government/news/uk-united-states-announce-partnership-on-science-of-ai-safety","evidenceUrl":"https://www.gov.uk/government/news/uk-united-states-announce-partnership-on-science-of-ai-safety","language":"en","accessedAt":"2026-10-27","reuse":{"attribution":"Department for Science, Innovation and Technology, “UK & United States announce partnership on science of AI safety”, 2024-04-02.","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-ab-uk-synna-text","title":"UK screening guidance on synthetic nucleic acids for users and providers","authors":["GOV.UK (Department for Science, Innovation and Technology)"],"publisher":"GOV.UK (Department for Science, Innovation and Technology)","publicationDate":null,"version":"Retrieved for the candidate pack on 2026-10-28 (SHA-256 d12116062f1c3dbb8a384593927c08cf40ee37289dc82cf6011540b63bcb7dd8). The HTML attachment as served on 2026-10-28. Its screening methodology sections and its list of indicators of a suspicious transaction were read but are not cited. Retrieved again at admission on 2026-10-30 (SHA-256 52c2f4a8bfc22b7819ee11216d08966df73f980a87b187b9c4b7bd868a544262).","canonicalUrl":"https://www.gov.uk/government/publications/uk-screening-guidance-on-synthetic-nucleic-acids/uk-screening-guidance-on-synthetic-nucleic-acids-for-users-and-providers","evidenceUrl":"https://www.gov.uk/government/publications/uk-screening-guidance-on-synthetic-nucleic-acids/uk-screening-guidance-on-synthetic-nucleic-acids-for-users-and-providers","language":"en","accessedAt":"2026-10-28","reuse":{"attribution":"GOV.UK (Department for Science, Innovation and Technology), “UK screening guidance on synthetic nucleic acids for users and providers”.","licence":"No open redistribution licence asserted","note":"Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution."}}],"evidence":[{"id":"evidence-good-discusses-explosion-1965","sourceRole":"primary","support":"supports","locator":"Section 2, printed page 33 (scan page 3)","passage":"an ultraintelligent machine could design even better machines","passageType":"quote","verification":"machine_checked","checkedAt":"2026-09-17","note":"Source passage inspected by an automated research assistant. Human editorial review is pending."},{"id":"evidence-issued-rsp-3-2","sourceRole":"primary","support":"supports","locator":"PDF p. 1, cover: version and effective-date lines","passage":"Version 3.2 Effective April 29, 2026","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-03","note":"Verbatim excerpt checked by an automated research assistant on 2026-10-03 against the inspected PDF, byte-identical to the Internet Archive capture of 2026-05-05; it replaces an earlier attributed paraphrase. Human editorial review is pending."},{"id":"evidence-lineage-ii-yudkowsky-agenda","sourceRole":"primary","support":"supports","locator":"PDF p. 1, author block","passage":"The author block includes Eliezer Yudkowsky.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-25","note":"Direct source inspection; machine check only. Authorship only. Work dated to 2016 by the publisher announcement; retrieved PDF revision is unknown."},{"id":"evidence-xm-llama-commercial","sourceRole":"primary","support":"supports","locator":"Takeaways and opening paragraph describing research and commercial availability.","passage":"The announcement describes Llama 2 as available free for research and commercial use.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-29","note":"Passage checked by machine against the identified source expression. No human editorial acceptance or independent verification of implementation is implied."},{"id":"evidence-cn-qwen3-license","sourceRole":"primary","support":"supports","locator":"引言, paragraph beginning 我们开源了两个 MoE 模型的权重","passage":"Two MoE and six dense models are listed as released under Apache 2.0.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-09-29","note":"English paraphrase translated by machine from the cited original Chinese passage; no human translation review. Exploratory seed, not editorially accepted. Translation scope and ambiguities are retained in the claim qualification."},{"id":"evidence-provchange-fsf-v3-0-security-level-adjustment-later","sourceRole":"primary","support":"supports","locator":"PDF p. 9 (printed p. 9), §2.2 Misuse Critical Capability Levels, second paragraph, second sentence","passage":"FSF v3.0 states that an adjustment of its recommended security levels may occur if, for example, a model does not possess capabilities meaningfully different from other publicly available models that have weaker security applied, in which case the marginal benefit of higher security is limited.","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-v3-0-security-level-adjustment-earlier","sourceRole":"primary","support":"supports","locator":"PDF p. 5 (printed p. 5), Misuse Critical Capability Levels, paragraph beginning “These recommended security levels”, second sentence","passage":"FSF v2.0 states that an adjustment of its recommended security levels may occur if, for example, a model does not possess capabilities meaningfully different from other widely available models that have not demonstrably caused or contributed to severe risks.","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-v3-1-deployment-mitigation-scope-later","sourceRole":"primary","support":"supports","locator":"PDF p. 7 (printed p. 7), §1.3.5 Risk Acceptance Determination, misuse and ML R&D bullets, first sub-bullets, final sentences","passage":"FSF v3.1 states, of its misuse risk acceptance criterion on deployment mitigations, that this is required only for external deployment, not internal deployment or further development.","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-v3-1-deployment-mitigation-scope-earlier","sourceRole":"primary","support":"supports","locator":"PDF p. 7 (printed p. 7), §1.6 Summary of Risk Acceptance Criteria, misuse and ML R&D bullets, first sub-bullets, final sentences","passage":"FSF v3.0 states, of its misuse risk acceptance criterion on deployment mitigations, that this is required only for external deployment, not further development.","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-seoul-amazon-1","sourceRole":"primary","support":"supports","locator":"Opening sentence and first list of organisations (sixteen names)","passage":"The text opens by stating that the UK and Republic of Korea governments announced that the listed organisations have agreed to the commitments, and the list includes “Amazon”.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-04","note":"Editorial paraphrase, re-checked on 2026-10-04 by the integrating session against its own direct fetch of the cited expression. The list is identical in the Internet Archive capture of 2024-05-21. The integrating session wrote or selected this passage from its own reading of the source; it does not repeat a passage from the week 4b research pass. Human editorial review is pending."},{"id":"evidence-ts-c1-rosenfelder-1992","sourceRole":"primary","support":"supports","locator":"Message body, opening sentence","passage":"For those who still think the Turing Test is a sufficient test for intelligence, there is food for thought","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Checked against the archive member on 2026-10-06."},{"id":"evidence-ts-a4-clemmensen-1996-08-30-context","sourceRole":"primary","support":"supports","locator":"Message body, reply to the quoted question \"what would we produce for them?\"","passage":"I believe that an SI will be able to design and build whatever automated tools it needs in just this way, in a matter of a few days or weeks.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Read in the archive copy retrieved 2026-10-06."},{"id":"evidence-ts-b2-yudkowsky-2002-aibox-context","sourceRole":"primary","support":"supports","locator":"Message body, first unquoted paragraph","passage":"I'll pretend to be a brain in a box. You pretend to be the experimenter. I'll try to persuade you to let me out.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Checked against the archive HTML on 2026-10-06 (hash matched the 1996–2022 candidate search)."},{"id":"evidence-ts-b1-corrigibility","sourceRole":"primary","support":"supports","locator":"p. 1, footnote 2 (utility-function symbols elided)","passage":"this deception could lead an agent to maximize [...] only until it is powerful enough to avoid correction by its programmers [...] Bostrom (2014) refers to this as a “treacherous turn”.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Checked against the text layer on 2026-10-06; utility-function symbols elided. On 2026-10-30, by the maintainer’s decision of that day, the quotation was corrected to the archive capture of 2015-03-22 (quotation marks and apostrophes as in the capture), which a cross-family archive check read on 2026-10-28; the proposition is unchanged."},{"id":"evidence-ts-a3-christiano-2018-takeoff","sourceRole":"primary","support":"supports","locator":"Section \"Reasons to expect fast takeoff\", subsection \"Recursive self-improvement\"","passage":"Before there is AI that is great at self-improvement there will be AI that is mediocre at self-improvement. Powerful AI can be used to develop better AI (amongst other things). This will lead to runaway growth. This on its own is not an argument for discontinuity","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Checked against the current page on 2026-10-06."},{"id":"evidence-yk-t08-d","sourceRole":"primary","support":"supports","locator":"Section B.1 (The distributional leap), item 13","passage":"Many alignment problems of superintelligence will not naturally appear at pre-dangerous, passively-safe levels of capability.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Candidate pack passage T08-d. The quotation (15 words) matched the retrieved expression (SHA-256 b36f38a6fd1020bfbd7b9d941e150ea68561242f3a41a0b8f2a56b853a9a6db3) when checked on 2026-10-06. It also matched the posting-day version 1.0.0. Pack version note: edited post, current version 1.3.0; posting-day check pending."},{"id":"evidence-yk-t10-c","sourceRole":"primary","support":"supports","locator":"§3 'Some Defenses of a Model of Hard Takeoff', paragraph beginning 'But why believe that?', printed p. 31","passage":"…the evolutionary history of hominid intelligence doesn’t show any signs of diminishing returns—there’s no sign that evolution took ten times as long to produce each successive marginal improvement of hominid brains.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-06","note":"Candidate pack passage T10-c. The quotation (31 words) matched the retrieved expression (SHA-256 1fbf547c811aadb7785d9d681cc05c557c6786055ec15af14ae680537a924f4a) when checked on 2026-10-06."},{"id":"evidence-ec-e05-29","sourceRole":"primary","support":"supports","locator":"Abstract (NBER abstract page)","passage":"there is an urgent need to advance our understanding of how it could reshape our economic models, institutions and policies","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-16","note":"Candidate pack passage E05-29 (research/economics-mainstream/esoteric). The quotation (20 words) matched the retrieved current copy (SHA-256 8fb8e191f7c3d57f9136f42e237a3b753aa23c2c4e7c09bd04df5b2568381666) in the pack's verification of 2026-10-15. Re-checked at admission on 2026-10-16 against a copy retrieved that day, whose SHA-256 differs from the recorded one.","contextOnly":{"reason":"GPT cross-family check 2026-10-14, accepted 2026-10-15","requires":"\"It\" is Transformative Artificial Intelligence, named in the preceding clause (\"As we approach Transformative Artificial Intelligence (TAI), …\"); the abstract does not define TAI."}},{"id":"evidence-ec-e05-30","sourceRole":"primary","support":"supports","locator":"Abstract (NBER abstract page)","passage":"nine Grand Challenges: economic growth, innovation, income distribution, decision-making power, geoeconomics, information flows, safety risks, human well-being, and transition dynamics","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-16","note":"Candidate pack passage E05-30 (research/economics-mainstream/esoteric). The quotation (20 words) matched the retrieved current copy (SHA-256 8fb8e191f7c3d57f9136f42e237a3b753aa23c2c4e7c09bd04df5b2568381666) in the pack's verification of 2026-10-15. Re-checked at admission on 2026-10-16 against a copy retrieved that day, whose SHA-256 differs from the recorded one.","contextOnly":{"reason":"Fresh-context audit (A2) judged that the quotation misleads when shown alone.","requires":"Do not read 'safety risks' as x-risk without the full text."}},{"id":"lg-ev-oai-pamela-vagata-founding-member-2015","sourceRole":"primary","support":"supports","locator":"Body, paragraph naming the research director, CTO and other founding members","passage":"The announcement lists Pamela Vagata among “the group’s other founding members”, described as research engineers and scientists.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “Pamela Vagata” 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-goo-parmar-google-research-2017","sourceRole":"primary","support":"supports","locator":"Page 1, author and affiliation block","passage":"Niki Parmar is listed under Google Research.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “Niki Parmar∗ Google Research” 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-oai-leike-departure-2024-1","sourceRole":"primary","support":"supports","locator":"Post text (first post of the thread)","passage":"Leike writes that the previous day was his last as head of alignment, superalignment lead and executive at OpenAI.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “Yesterday” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through X’s own syndication endpoint for the post’s text, which truncates long notes. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. A cross-family machine check (2026-10-16/17) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"x_syndication"},{"id":"lg-ev-oai-leike-departure-2024-2","sourceRole":"primary","support":"supports","locator":"Post text (second post of the thread)","passage":"Leike writes that he had long disagreed with OpenAI leadership about the company’s core priorities until they reached a breaking point.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-15","note":"Candidate pack passage, lab genealogy (research/lab-genealogy). The paraphrase’s verbatim anchor “core priorities” matched the retrieved text on 2026-10-15 and again in a fresh retrieval the same day. Read through X’s own syndication endpoint for the post’s text, which truncates long notes. Matching ignored differences of whitespace, case, quotation marks, apostrophes, dashes and Unicode compatibility forms, so the source’s typography can differ. The paraphrase is Selekoa’s; only the anchor is the source’s wording. A cross-family machine check (2026-10-16/17) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"x_syndication"},{"id":"ev-cmp-meta-aasf-v2-0-d6-discretion-1","sourceRole":"primary","support":"supports","locator":"PDF p. 15 (printed p. 15), §3.3 Risk Thresholds, Table 1, note below the Measures column, second sentence","passage":"Final deployment decisions may also consider additional factors beyond those covered in this Framework.","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. Human editorial review is pending."},{"id":"evidence-ap-daron-acemoglu-acemoglu-restrepo-2019-wrong-kind-of-ai","sourceRole":"primary","support":"supports","locator":"PDF text layer, first two pages: author name on the title page or first page","passage":"Daron Acemoglu","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-27","note":"Credited name checked by an automated research assistant on 2026-10-27 in a direct retrieval. Human editorial review is pending.","retrievalRoute":"direct"},{"id":"evidence-ap-daniel-rock-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":"Daniel Rock","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":"ev-cmp-rsp-2023-d7-internal-1","sourceRole":"primary","support":"supports","locator":"PDF p. 10 (printed p. 10), Procedural Commitments, numbered item 7 ('Responsible Scaling Officer'), second sentence","passage":"The edition says the Responsible Scaling Officer will report each quarter to the board and the LTBT on implementation status, noting any deficiencies in implementation.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-27","note":"Explicit paraphrase, written on 2026-10-27 by an automated research assistant after re-reading the passage in the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. The comparison quoted the passage; it is given here as a paraphrase to keep the edition within the quotation limit. Item 6 commits to sharing ASL evaluation results promptly with the board and LTBT; item 8 requires a non-compliance reporting policy allowing anonymous feedback. Human editorial review is pending."},{"id":"ev-cmp-faif-2025-d2-levels-1","sourceRole":"primary","support":"supports","locator":"PDF p. 6, 'Addressing Risks of Malicious Use' › '2. Risk Assessment', paragraph beginning 'Thresholds:', second sentence","passage":"Under “Thresholds” for malicious use, the edition states that the criterion xAI uses to accept the risk of deploying a system is an answer rate on restricted queries below one in twenty.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-27","note":"Explicit paraphrase, written on 2026-10-27 by an automated research assistant after re-reading the passage in the publisher copy, whose SHA-256 matches the copy retrieved on 2026-10-15. The comparison quoted the passage; it is given here as a paraphrase to keep the edition within the quotation limit. The restricted queries belong to an internal benchmark of biology and chemistry queries developed with SecureBio (same paragraph). The second numeric criterion, a MASK dishonesty rate, is recorded under “Risk domains covered” Other named domains. Human editorial review is pending."},{"id":"c2-ev-governance-aiia-aisc-signatory-china-mobile-1","sourceRole":"primary","support":"supports","locator":"Embedded image, list 首批签署《人工智能安全承诺》企业名单 (transcribed from the image; not in the page’s text layer)","passage":"Transcription from the image: the list of initial signatories (首批签署《人工智能安全承诺》企业名单, 17 names, ordered by initial letter) includes 中国移动通信集团有限公司.","passageType":"paraphrase","verification":"machine_checked","checkedAt":"2026-10-27","note":"Candidate pack passage, Chinese frontier-AI institutions, second collection (research/china-labs-2). Transcription of a name that appears only in the image embedded in the post (the list of initial signatories); it is not in the page’s text layer and was not machine-matched. The drafting pass and the same-family audit read the list visually, and the cross-family check (2026-10-27) read all 17 names and agreed. The paraphrase is Selekoa’s; the listed name is transcribed as printed. No English rendering is given; no human translation review. A cross-family machine check (2026-10-27, item K013) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"c2-ev-governance-aiia-aisc-signatory-china-mobile-2","sourceRole":"secondary","support":"supports","locator":"Final paragraph","passage":"商汤科技、稀宇科技、腾讯、电信AI、中国移动已完成承诺签署","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 (24 words, each CJK character counted as one) 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. English machine translation, not verified by a person: “SenseTime, MiniMax (Xiyu Technology), Tencent, Telecom AI and China Mobile have completed signing the commitments”. A cross-family machine check (2026-10-27, item K013) re-read the record’s sources and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-uk-us-aisi-mou-signed-1","sourceRole":"primary","support":"supports","locator":"Opening paragraph","passage":"The UK and US have signed (Monday 1 April 2024) a Memorandum of Understanding (MOU) which will see them work together to develop tests for the most advanced artificial intelligence (AI) models","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (32 words) matched the text retrieved for the candidate pack on 2026-10-27 and again at admission on 2026-10-28, in a fresh retrieval of the same address (the page’s bytes had changed). Read directly from www.gov.uk. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. Signatories named in the next sentence. A cross-family machine check (2026-10-28, item 62) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-fl-uk-us-aisi-mou-signed-2","sourceRole":"primary","support":"supports","locator":"Opening, second sentence","passage":"Signed by Technology Secretary Michelle Donelan and US Commerce Secretary Gina Raimondo","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-28","note":"Candidate pack passage, frameworks to law (research/frameworks-to-law). The quotation (12 words) matched the text retrieved for the candidate pack on 2026-10-27 and again at admission on 2026-10-28, in a fresh retrieval of the same address (the page’s bytes had changed). Read directly from www.gov.uk. Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. Public roles as stated in the release. A cross-family machine check (2026-10-28, item 62) answered the record’s question blind from fresh retrievals and agreed with the record. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-scr-uk-guidance-ai-in-background-1","sourceRole":"primary","support":"supports","locator":"Background section","passage":"Increased access to synthetic nucleic acids, coupled with developments in converging tools and technologies – such as AI – and an increase in knowledge and expertise, may exacerbate the risks of engineering biology.","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (31 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval (SHA-256 52c2f4a8bfc22b7819ee11216d08966df73f980a87b187b9c4b7bd868a544262; the bytes differ from the pack’s retrieval). Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. The preceding sentence of the guidance describes what could make it easier for hostile actors to obtain components of dangerous biological systems; it is not reproduced. No human editorial review.","retrievalRoute":"direct"},{"id":"evidence-ab-gov-eu-ai-act-recital110-cbrn-1","sourceRole":"primary","support":"supports","locator":"Recital 110","passage":"chemical, biological, radiological, and nuclear risks, such as the ways in which barriers to entry can be lowered","passageType":"quote","verification":"machine_checked","checkedAt":"2026-10-30","note":"Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (18 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 recital continues \"including for weapons development, design acquisition, or use\" and lists further risks (offensive cyber capabilities, tool use, self-replication, bias, disinformation, chain reactions). A cross-family machine check (2026-10-30, item 83) re-read the record’s sources and found full support. No human editorial review.","retrievalRoute":"direct"}],"at":{"entities":[0,1,5,19,21,23,24,27,40,47,50,51,52,56,74,105,157,168,180,185,200,233,252,276,286,312,322,327,333,338,339,342,343,347,357,359,491,548,564,589,612,754,762,848,915,936,940,1030,1210],"claims":[24,63,245,352,376,595,596,700,829,861,916,942,953,1043,1053,1564,1576,1583,1689,2157,2175,2316,2490,2675,2753,3025,3309,3332],"sources":[0,10,11,35,41,45,46,47,51,87,104,158,168,180,239,278,287,316,327,332,338,457,500,517,562,750,751,778,787,864,1053],"evidence":[25,66,266,375,399,669,670,671,672,879,1040,1075,1133,1164,1176,1271,1282,1957,1958,1973,1980,2134,2135,2801,2820,2963,3137,3356,3471,3472,3928,3929,4540,4579]},"version":"1lg47t7"}