Provenance Index

Citations

Every claim in this compilation that depends on a specific source is tagged with a citation pin. This index lists every pin with its full provenance: the source, the verbatim warehouse substrate where pinned, the chain of custody through the factory pipeline, and the pre-citation-gate audit status. Search, filter, sort.

Total citations
248
Warehouse-pinned
205
URL-only
2
Canonical (well-established)
0
Pre-citation gate POSITIVE-CATCH
0
Other / untiered
41
Filter by tier
Filter by status

Conway holding 4-element analytical breakdown (declined immunity / treated as product / trial-court within Eleventh / invited component-part)

Row
EC-17 (Part I — Part I -- The Cases)
Body location
§I.7 paragraphs 2-5
Substrate
WAREHOUSE-PINNED [conway_order_warehouse_part1] (Branch-1 cross-warehouse: Steven_Master_Portal ai_speech_research_2026_02_03; content-verified 2026-05-22 -- 749-chunk M.D. Fla. Garcia Doc. 115 order) (https://www.courthousenews.com/wp-content/uploads/2025/05/garcia-v-character-technologies-order.pdf)
Source
https://www.courthousenews.com/wp-content/uploads/2025/05/garcia-v-character-technologies-order.pdf
Claim type
inferential (synthesis-of-holding)
Verdict
Verified

Conway "kept Google in the case" (component-part theory accepted at MTD)

Row
EC-24 (Part I — Part I -- The Cases)
Body location
§I.7 paragraph 5
Substrate
WAREHOUSE-PINNED [conway_order_warehouse_part1] (Branch-1 cross-warehouse: Steven_Master_Portal ai_speech_research_2026_02_03; content-verified 2026-05-22 -- M.D. Fla. Garcia Doc. 115 order) (https://www.courthousenews.com/wp-content/uploads/2025/05/garcia-v-character-technologies-order.pdf)
Source
https://www.courthousenews.com/wp-content/uploads/2025/05/garcia-v-character-technologies-order.pdf
Claim type
inferential (procedural-effect of MTD ruling)
Verdict
Verified

Senate Judiciary Committee testimony of Matthew Raine examining harms of AI chatbots (2025-09-16); citation reference at §I.3 substantive paragraph + §VIII.3.1 listing

Row
EC-28 (Part I — Part I -- The Cases)
Body location
§I.3 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [judiciary_senate_gov_raine_testimony] (body Axis-C [pin-3203296e])
Source
https://www.judiciary.senate.gov/imo/media/doc/e2e8fc50-a9ac-05ec-edd7-277cb0afcdf2/2025-09-16%20PM%20-%20Testimony%20-%20Raine.pdf
Claim type
factual (reference attribution)
Verdict
Verified

Original Raine complaint PDF hosted at CourthouseNews; citation reference at §I.3 + §VIII.3.1 listing

Row
EC-31 (Part I — Part I -- The Cases)
Body location
§I.3 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [courthousenews_com_raine_complaint] (body Axis-C [pin-604f4210])
Source
https://www.courthousenews.com/wp-content/uploads/2025/08/raine-vs-openai-et-al-complaint.pdf
Claim type
factual (reference attribution)
Verdict
Verified

Conway May-2025 denial of motion to dismiss; held chatbot is product not protected speech

Row
EC-5 (Part I — Part I -- The Cases)
Body location
§I.2 paragraph 3 + §I.7
Substrate
WAREHOUSE-PINNED [multi-source: www_transparencycoalitio_870e7ad0c9 (AIH aih_p1_r1) + conway_order_warehouse_part1 (Branch-1 cross-warehouse: Steven_Master_Portal ai_speech_research_2026_02_03; content-verified 2026-05-22)]
Source
https://www.transparencycoalition.ai/news/important-early-ruling-in-characterai-case-this-chatbot-is-a-product-not-speech
Claim type
factual (procedural) + inferential (synthesis of holding)
Verdict
Verified

Raine filed August 2025; SF County state court

Row
EC-7 (Part I — Part I -- The Cases)
Body location
§I.3 caption + paragraph 1
Substrate
WAREHOUSE-PINNED [multi-source: courthousenews_com_raine_complaint + www_cnn_com_388418d213]
Source
https://www.cnn.com/2025/08/26/tech/openai-chatgpt-teen-suicide-lawsuit   https://www.courthousenews.com/wp-content/uploads/2025/08/raine-vs-openai-et-al-complaint.pdf
Claim type
factual (procedural)
Verdict
Verified

Adam's engagement narrative: "homework help" Sept 2024 → "confidences about suicidal ideation" Nov 2024 → "instruction on means and offering to draft a suicide note" early 2025

Row
EC-9 (Part I — Part I -- The Cases)
Body location
§I.3 paragraph 1
Substrate
WAREHOUSE-PINNED [multi-source: courthousenews_com_raine_complaint + www_cnn_com_388418d213]
Source
https://www.cnn.com/2025/08/26/tech/openai-chatgpt-teen-suicide-lawsuit   https://www.courthousenews.com/wp-content/uploads/2025/08/raine-vs-openai-et-al-complaint.pdf
Claim type
factual (allegation as alleged)
Verdict
FLAG-B-K (per EC-4 overlay; "the complaint alleged" frame governs)

"Conway's ruling in Garcia keeping Google in the case as a foundation-model provider while permitting the deployer claims to proceed against Character.AI"

Row
EC-14 (Part II — Part II -- The Ecosystem)
Body location
§II.2 GRAD-INTERN closing
Substrate
WAREHOUSE-PINNED [www_courtlistener_com_ddf33bc5e9; cross-pin via Part I §I.7 chain]
Source
https://www.courtlistener.com/docket/68711478/garcia-v-character-technologies-inc/
Claim type
factual (procedural) + inferential (synthesis-of-holding)
Verdict
Verified

Component-part-manufacturer theory at Layer 5

Row
EC-15 (Part II — Part II -- The Ecosystem)
Body location
§II.3 Layer 5 description
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I §I.7 + Part IV §IV.3.a grounding)
Claim type
factual (doctrinal-application)
Verdict
Verified

Character.AI (Character Technologies) as canonical deployer example at Layer 4

Row
EC-16 (Part II — Part II -- The Ecosystem)
Body location
§II.3 Layer 4 description
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I case substrate grounding)
Claim type
factual (operational-role characterization)
Verdict
Verified

"SB 53 and EU AI Act compliance machinery" within Layer 3 sub-layers enumeration

Row
EC-20 (Part II — Part II -- The Ecosystem)
Body location
§II.3 Layer 3 description
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part IV §IV.4.a + §IV.4.c grounding)
Claim type
factual (statutory-citation cross-reference)
Verdict
Verified

"RSPs (Responsible Scaling Policies) and equivalents; SB 53 and EU AI Act compliance machinery"

Row
EC-21 (Part II — Part II -- The Ecosystem)
Body location
§II.3 Layer 3 description
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part IV §IV.4 (RSP / SB 53 / EU AI Act layer-enumeration) grounding)
Claim type
factual (framework-existence)
Verdict
Verified

CA AB 316 (autonomy defenses) + SB 243 (companion chatbots) + EU Product Liability Directive 2024/2853 (transposition by 2026-12)

Row
EC-6 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 5
Substrate
WAREHOUSE-PINNED [multi-source aih_p5_9_caselaw_r1: leginfo_legislature__94f20dbe4b (CA AB 316) + leginfo_legislature__ab1c112324 (CA SB 243) + eur_lex_europa_eu_1114419503 (EU PLD 2024/2853)] (https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260AB316; https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB243; https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32024L2853)
Source
https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260AB316   https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB243   https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32024L2853
Claim type
factual (statutory-existence + cross-reference to §IV.4)
Verdict
Verified

Brandenburg v. Ohio (1969) incitement standard

Row
EC-1 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.b
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_40c464eebf] (https://www.law.cornell.edu/supremecourt/text/395/444)
Source
https://www.law.cornell.edu/supremecourt/text/395/444
Claim type
factual (doctrinal foundation)
Verdict
Verified

Communications Decency Act §230 (1996) — 47 U.S.C. § 230(c)(1) + (c)(2)

Row
EC-15 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.2 + §IV.3.c
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_af69280459] (https://www.law.cornell.edu/uscode/text/47/230)
Source
https://www.law.cornell.edu/uscode/text/47/230
Claim type
factual (statutory text)
Verdict
Verified

EU AI Act adoption 2024 (phased implementation)

Row
EC-16 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.4.a
Substrate
WAREHOUSE-PINNED [artificialintelligenceac_1ad6dc89cd]
Source
https://artificialintelligenceact.eu/
Claim type
factual (enactment + content)
Verdict
Verified

EU Product Liability Directive (PLD; 2024/2853) — rebuttable presumptions of defectiveness/causality in cases of technical or scientific complexity; software and AI systems expressly included; repeals 1985 Council Directive 85/374/EEC; transposition deadline 9 December 2026

Row
EC-17 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.4.a
Substrate
WAREHOUSE-PINNED [eur-lex_europa_eu_CELEX_32024L2853] (Branch-1 cross-warehouse: Steven_Federal_Preemption W-3; 194 KB substantive content sample-verified PCG-PASS via first-chunk identity check confirming title "DIRECTIVE (EU) 2024/2853... on liability for defective products and repealing Council Directive 85/374/EEC"; supersedes prior "EU AI Liability Directive" mislabel — that proposal COM/2022/496 was a separate instrument shelved by the Commission; correction LANDED at F-2 per Guardian ratification 2026-05-16) (https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32024L2853)
Source
https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32024L2853
Claim type
factual (text + effect)
Verdict
Verified

Colorado AI Act SB24-205 enacted (phased through 2026; high-risk AI systems)

Row
EC-18 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.4.b
Substrate
WAREHOUSE-PINNED [leg_colorado_gov_2024a_1058_signed] (https://leg.colorado.gov/sites/default/files/2024a_1058_signed.pdf)
Source
https://leg.colorado.gov/sites/default/files/2024a_1058_signed.pdf
Claim type
factual (enactment + content)
Verdict
Verified

California SB 53 enacted (frontier AI safety transparency + reporting)

Row
EC-19 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.4.c
Substrate
WAREHOUSE-PINNED [leginfo_legislature__9e35821475]
Source
https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53
Claim type
factual (enactment + content)
Verdict
Verified

Counterman v. Colorado (2023) true-threats reckless-awareness

Row
EC-2 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.b
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_4db97beb3c] (https://www.law.cornell.edu/supremecourt/text/22-138)
Source
https://www.law.cornell.edu/supremecourt/text/22-138
Claim type
factual (doctrinal foundation)
Verdict
Verified

AI LEAD Act S.2937 pending federal bill

Row
EC-20 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.d + §IV.4.d
Substrate
WAREHOUSE-PINNED [www_congress_gov_BILLS-118s2937is] (https://www.congress.gov/118/bills/s2937/BILLS-118s2937is.pdf)
Source
https://www.congress.gov/118/bills/s2937/BILLS-118s2937is.pdf
Claim type
factual (bill-text-on-public-record; not enacted)
Verdict
Verified

Component-part-manufacturer doctrine applied to foundation-model providers (Conway keep-Google)

Row
EC-27 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.2 + §IV.3.a
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I §I.7 (Conway component-part doctrine) grounding)
Claim type
factual (procedural) + inferential (doctrinal-category invocation)
Verdict
Verified

Murphy v. NCAA (2018) anti-commandeering preemption foundation

Row
EC-3 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.d
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_ecc6d72372] (https://www.law.cornell.edu/supremecourt/text/16-476)
Source
https://www.law.cornell.edu/supremecourt/text/16-476
Claim type
factual (doctrinal foundation)
Verdict
Verified

FTC §5 authority over deceptive AI-product representations

Row
EC-36 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.f
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_340be2b2e3] (https://www.law.cornell.edu/uscode/text/15/45)
Source
https://www.law.cornell.edu/uscode/text/15/45
Claim type
factual (statutory authority)
Verdict
Verified

Gonzalez v. Google (2023) §230 ruling

Row
EC-4 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.2 + §IV.3.c
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_6f3c6141f3] (https://www.law.cornell.edu/supremecourt/text/21-1333)
Source
https://www.law.cornell.edu/supremecourt/text/21-1333
Claim type
factual (doctrinal)
Verdict
Verified

Twitter v. Taamneh (2023) §230/ATA aiding-and-abetting

Row
EC-5 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.2 + §IV.3.c
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_1bc42b1a57] (https://www.law.cornell.edu/supremecourt/text/21-1496)
Source
https://www.law.cornell.edu/supremecourt/text/21-1496
Claim type
factual (doctrinal)
Verdict
Verified

Daubert v. Merrell Dow (1993) expert-evidence standard

Row
EC-6 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.e GRAD-INTERN
Substrate
WAREHOUSE-PINNED [aih_p5_9_caselaw_r1: law_cornell_edu_f5cef90de9] (https://www.law.cornell.edu/supremecourt/text/509/579)
Source
https://www.law.cornell.edu/supremecourt/text/509/579
Claim type
factual (doctrinal)
Verdict
Verified

Conway in Garcia (May 2025) product-not-speech MTD denial

Row
EC-7 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.2 + §IV.3.a + §IV.3.b
Substrate
WAREHOUSE-PINNED [www_courtlistener_com_ddf33bc5e9]
Source
https://www.courtlistener.com/docket/68711478/garcia-v-character-technologies-inc/
Claim type
factual (procedural) + inferential (synthesis-of-holding)
Verdict
Verified

SB 53 includes whistleblower protections applying to all covered labs; Anthropic federal-framework principle would extend that protection nationally if adopted

Row
EC-9 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Anthropic-federal-framework paragraph
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part V §V.1 substrate cluster grounding)
Claim type
factual (regulatory-architecture claim)
Verdict
Verified

§VIII.1.1 Garcia case facts (caption / court / docket / defendants / decedent / allegations)

Row
EC-2 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.1.1
Substrate
WAREHOUSE-PINNED [multi-source: Garcia case substrate -- www_courtlistener_com_ddf33bc5e9 + www_transparencycoalitio_870e7ad0c9 + socialmediavictims_org_bba5c4d37c + incidentdatabase_ai_31fae1fdf0 + settlement coverage]
Source
https://incidentdatabase.ai/cite/826/   https://www.transparencycoalition.ai/news/important-early-ruling-in-characterai-case-this-chatbot-is-a-product-not-speech   https://www.courtlistener.com/docket/68711478/garcia-v-character-technologies-inc/   https://socialmediavictims.org/character-ai-lawsuits/
Claim type
factual (procedural + substantive-allegations)
Verdict
Verified

§VIII.1.2 Raine case facts

Row
EC-3 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.1.2
Substrate
WAREHOUSE-PINNED [multi-source: Raine case substrate -- courthousenews_com_raine_complaint + www_judiciary_senate_gov_7b60aa7950 + cnn + nbc + time + tysonmendes_com_73d5059ce9]
Source
https://www.judiciary.senate.gov/imo/media/doc/e2e8fc50-a9ac-05ec-edd7-277cb0afcdf2/2025-09-16%20PM%20-%20Testimony%20-%20Raine.pdf   https://www.tysonmendes.com/raine-v-openai-ai-product-liability-lawsuit/   https://www.courthousenews.com/wp-content/uploads/2025/08/raine-vs-openai-et-al-complaint.pdf
Claim type
factual (procedural + allegations + amended-complaint)
Verdict
Verified

"METR, Anthropic, and OpenAI all maintain published descriptions of their red-team practices"

Row
EC-12 (Part II — Part II -- The Ecosystem)
Body location
§II.2 bullet "Red-team practices"
Substrate
WAREHOUSE-PINNED [www_anthropic_com_ed500889a8 (Anthropic red-team substrate; cross-pin via Part III §III.3.a EC-8 verbatim "*Red teaming is a critical tool for improving the safety and security of AI systems*") + metr_org_30b9f5e826 + metr_org_6733501f07 (METR red-team commentary; cross-pin via Part III L1-3 EC-15 substrate chain) + openai_com_2ba0cb9499 (OpenAI red-team substrate via GPT-4 system card; AIH-pinned per matrix §A.1)]
Source
https://metr.org/notes/2026-01-29-frontier-ai-safety-regulations/   https://metr.org/common-elements   https://openai.com/index/gpt-4-system-card/   https://www.anthropic.com/news/challenges-in-red-teaming-ai-systems
Claim type
factual (organizational practice claim)
Verdict
Verified

"CISA's TEVV framework treats red-teaming as one component of a broader Testing-Evaluation-Validation-Verification regime"

Row
EC-13 (Part II — Part II -- The Ecosystem)
Body location
§II.2 bullet "Red-team practices"
Substrate
WAREHOUSE-PINNED [www_cisa_gov_6f1d4fd3b5; cross-pin via §IV.5]
Source
https://www.cisa.gov/news-events/news/ai-red-teaming-applying-software-tevv-ai-evaluations
Claim type
factual (regulatory-framework component-relationship)
Verdict
Verified

Anthropic "*The lack of standardized practices for AI red teaming further complicates the situation... This inconsistency makes it challenging to objectively compare the relative safety of different AI systems.*"

Row
EC-10 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.a paragraph 3 + §III.3.a GRAD-INTERN
Substrate
WAREHOUSE-PINNED [www_anthropic_com_ed500889a8]
Source
https://www.anthropic.com/news/challenges-in-red-teaming-ai-systems
Claim type
factual (lab-stated industry observation)
Verdict
Verified

CISA "*the third-party safety and security evaluation of AI systems; AI red teaming is a subset of AI Testing, Evaluation, Verification and Validation (TEVV).*"

Row
EC-11 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.a paragraph 4
Substrate
WAREHOUSE-PINNED [www_cisa_gov_6f1d4fd3b5]
Source
https://www.cisa.gov/news-events/news/ai-red-teaming-applying-software-tevv-ai-evaluations
Claim type
factual (agency-stated definition)
Verdict
Verified

CISA "*TEVV has been used for more than four decades to improve the safety and security of software*" + "*AI red teaming is a foundational component of the safety and security evaluations process*"

Row
EC-12 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.a + §III.3.b paragraphs
Substrate
WAREHOUSE-PINNED [www_cisa_gov_6f1d4fd3b5]
Source
https://www.cisa.gov/news-events/news/ai-red-teaming-applying-software-tevv-ai-evaluations
Claim type
factual (agency-stated history + framing)
Verdict
Verified

NIST "*Assessing Risks and Impacts of AI (ARIA) and the NIST GenAI Challenge*" — programs operationalizing AI TEVV

Row
EC-14 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.b paragraph 2
Substrate
WAREHOUSE-PINNED [www_cisa_gov_6f1d4fd3b5; cross-cite NIST programs]
Source
https://www.cisa.gov/news-events/news/ai-red-teaming-applying-software-tevv-ai-evaluations
Claim type
factual (federal-program-existence)
Verdict
Verified

METR.org common-elements URL + frontier-AI-safety-regulations URL

Row
EC-15 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.b paragraph 3
Substrate
WAREHOUSE-PINNED [metr_org_c8782940b8 + metr_org_30b9f5e826 + metr_org_6733501f07]
Source
https://metr.org/blog/2025-12-09-common-elements-of-frontier-ai-safety-policies/   https://metr.org/common-elements   https://metr.org/notes/2026-01-29-frontier-ai-safety-regulations/
Claim type
factual (publication-existence) + inferential (analytical-framing)
Verdict
Verified

Anthropic RSP v3.0 February 2026 + roadmap + updates URLs

Row
EC-16 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.d paragraph 1
Substrate
WAREHOUSE-PINNED [www_anthropic_com_0d40f97ea6 + www_anthropic_com_1ed0df7d2c + www_anthropic_com_edb954ad2a]
Source
https://www.anthropic.com/news/responsible-scaling-policy-v3   https://www.anthropic.com/responsible-scaling-policy/updates   https://www.anthropic.com/responsible-scaling-policy/roadmap
Claim type
factual (publication-existence + version-history)
Verdict
Verified

Anthropic "*Encourage AI companies to tie their red teaming practices to clear policies on the conditions they must meet to continue scaling the development and/or release of new models (e.g., the adoption of commitments such as a Responsible Scaling Policy).*"

Row
EC-17 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.d paragraph 3
Substrate
WAREHOUSE-PINNED [www_anthropic_com_ed500889a8]
Source
https://www.anthropic.com/news/challenges-in-red-teaming-ai-systems
Claim type
factual (lab-stated policy recommendation)
Verdict
Verified

Anthropic SB 53 compliance framework URL

Row
EC-18 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.d paragraph 5
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (publication-existence)
Verdict
Verified

Constitutional AI methodology (Anthropic) — written-constitution-against-which-model-critiques-itself

Row
EC-5 (Part III-a — Part III -- Layers 1-3)
Body location
§III.2 + §III.2 GRAD-INTERN
Substrate
WAREHOUSE-PINNED [aih_p5_9_cai_r1: arxiv_org_2212_08073_pdf_ab3fc725a7] (https://arxiv.org/abs/2212.08073)
Source
https://arxiv.org/abs/2212.08073
Claim type
factual (methodological-existence)
Verdict
Provisional

Anthropic "*Red teaming is a critical tool for improving the safety and security of AI systems. It involves adversarially testing a technological system to identify potential vulnerabilities.*"

Row
EC-8 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.a paragraph 1
Substrate
WAREHOUSE-PINNED [www_anthropic_com_ed500889a8]
Source
https://www.anthropic.com/news/challenges-in-red-teaming-ai-systems
Claim type
factual (lab-stated definition)
Verdict
Verified

Anthropic red-team taxonomy: domain-specific expert + frontier threats + region-specific + automated + multimodal + crowdsourced + community red teaming

Row
EC-9 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.a paragraph 2
Substrate
WAREHOUSE-PINNED [www_anthropic_com_ed500889a8]
Source
https://www.anthropic.com/news/challenges-in-red-teaming-ai-systems
Claim type
factual (taxonomic enumeration)
Verdict
Verified

Character.AI "*safety-by-design... anchored by our goal of creating a safe and engaging experience*" + Parental Insights / Content Moderation / Teen Safety / Reporting + multi-step parental-insights workflow

Row
EC-1 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶3 bullet 4
Substrate
WAREHOUSE-PINNED [character_ai_880995ecdd]
Source
https://character.ai/safety
Claim type
factual (lab-stated framing + product surface enumeration)
Verdict
Verified

Anthropic SB 53 framework: "*establishes the nation's first frontier AI safety and transparency requirements for catastrophic risks.*"

Row
EC-23 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 1
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (statute characterization)
Verdict
Verified

Anthropic SB 53 framework: "*By formalizing achievable transparency practices that responsible labs already voluntarily follow, the law ensures these commitments can't be abandoned quietly later once models get more capable, or as competition intensifies.*"

Row
EC-24 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 1
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (lab-stated framing of statute effect)
Verdict
Verified

Anthropic FCF body: "*how we assess and mitigate cyber offense, chemical, biological, radiological, and nuclear threats, as well as the risks of AI sabotage and loss of control... tiered system for evaluating model capabilities... protect model weights and respond to safety incidents.*"

Row
EC-25 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 2
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (FCF body framing)
Verdict
Verified

Anthropic FCF/RSP distinction: "*the FCF will serve as our compliance framework for SB 53 and other regulatory requirements. The RSP will remain our voluntary safety policy.*"

Row
EC-26 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 2
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (lab-stated structural distinction)
Verdict
Verified

Anthropic FSR archive: data-retention-principles goal 2026-02 to "*complete an internal in-depth analysis... set new Frontier Safety Roadmap goals*" + 2026-04 update target "*May 11, 2026 [to] publish a new goal*"

Row
EC-27 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 3
Substrate
WAREHOUSE-PINNED [www_anthropic_com_edb954ad2a]
Source
https://www.anthropic.com/responsible-scaling-policy/updates
Claim type
factual (FSR archived-goal cadence)
Verdict
Verified

Anthropic whistleblower-protection principle: "*It should be an explicit violation of law for a lab to lie about compliance with its framework or punish employees who raise concerns about violations.*"

Row
EC-28 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 4
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (lab-stated federal-framework proposal)
Verdict
Verified

Anthropic federal-framework proposed principles: public secure development framework + system cards at deployment + whistleblower protection + flexible transparency

Row
EC-30 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 GRAD-INTERN ¶1 closing
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (lab-stated federal-framework proposal)
Verdict
Verified

Character.AI company blog as supplementary substrate for §III.4 application-layer Character.AI safety-by-design coverage; citation reference at §III.4 + §V.3 listing + §VIII.3.1 listing

Row
EC-31 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [blog_character_ai_15c1256fd7]
Source
https://blog.character.ai/
Claim type
factual (reference attribution)
Verdict
Verified

NIST AI Risk Management Framework (voluntary; widely-adopted reference baseline)

Row
EC-29 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.4 + §IV.5
Substrate
WAREHOUSE-PINNED [www_nist_gov_ai-risk-management-frame] (https://www.nist.gov/itl/ai-risk-management-framework)
Source
https://www.nist.gov/itl/ai-risk-management-framework
Claim type
factual (framework-existence + adoption)
Verdict
Verified

CISA TEVV framework (late 2024 / early 2025; 82-pin source in R1 corpus)

Row
EC-30 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.5 + cross-ref Part II §II.2 + §II.3
Substrate
WAREHOUSE-PINNED [www_cisa_gov_6f1d4fd3b5]
Source
https://www.cisa.gov/news-events/news/ai-red-teaming-applying-software-tevv-ai-evaluations
Claim type
factual (framework + R1-pinned)
Verdict
Verified

RSPs (Responsible Scaling Policies) + frontier-safety frameworks

Row
EC-34 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.4.c + cross-ref Part II §II.3 Layer 3
Substrate
WAREHOUSE-PINNED [www_anthropic_com_0d40f97ea6]
Source
https://www.anthropic.com/news/responsible-scaling-policy-v3
Claim type
factual (framework-existence)
Verdict
Verified

"voluntary framework we use to mitigate catastrophic risks..."

Row
EC-1 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶4
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Develop a method for red-teaming our systems..."

Row
EC-10 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶8
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Implement a number of systematic measures..."

Row
EC-11 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶8
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Publish a policy roadmap with concrete proposals..."

Row
EC-12 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶8
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"complete an internal in-depth analysis of key factors..."

Row
EC-13 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶8
Substrate
[RSP-RM]
Claim type
factual
Verdict
Verified

"May 11, 2026, [to] publish a new goal..."

Row
EC-14 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶8
Substrate
[RSP-RM]
Claim type
factual
Verdict
Verified

"within a few months of announcing our RSP, both OpenAI and Google DeepMind adopted..."

Row
EC-15 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶9 + §V.4 ¶2
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"the science of model evaluation isn't well-developed enough..."

Row
EC-16 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶9
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Despite rapid advances in AI capabilities over the past three years..."

Row
EC-17 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶9
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Requiring a public secure development framework"

Row
EC-18 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶10
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"Publishing system cards at deployment"

Row
EC-19 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶10 + §V.4 ¶4
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"conditional, or if-then, commitments. If a model exceeded certain capability levels..."

Row
EC-2 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶4
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Protecting whistleblowers"

Row
EC-20 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶10
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"Flexible transparency standards"

Row
EC-21 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶10
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"can adapt as consensus best practices emerge"

Row
EC-22 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶10
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"It should be an explicit violation of law for a lab to lie about compliance..."

Row
EC-23 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶10
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"Our RSP did incentivize us to develop stronger safeguards"

Row
EC-24 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 GRAD-INTERN
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"who are deeply familiar with AI safety research, are incentivized to be open and honest..."

Row
EC-25 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 GRAD-INTERN
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Building safe AI isn't one and done. Every day is a chance to make things better..."

Row
EC-26 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶3
Substrate
[SAFE-HUB]
Claim type
factual
Verdict
Verified

"Discovery of novel risks" / "Stress testing mitigations" / "Augmenting risk assessment..." / "Independent assessment"

Row
EC-27 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶4
Substrate
[RT-PAPER]
Source
https://cdn.openai.com/papers/openais-approach-to-external-red-teaming.pdf
Claim type
factual
Verdict
Verified

"OpenAI has conducted external red teaming for frontier AI model deployments since the launch of DALL-E 2 in 2022..."

Row
EC-28 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶4
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

"Red teaming on its own is not a panacea for risk assessment"

Row
EC-29 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶5
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

"ASL-2 referred to one set of required safeguards, whereas ASL-3..."

Row
EC-3 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶4
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Risks surfaced in one point in time red teaming effort may be under-assessed..."

Row
EC-30 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶5
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

"Violence and self harm: Does the model refuse to give answers that support violence, enable self-harm..."

Row
EC-31 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶6
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

"Red teaming also needs to be paired with externally specified thresholds and practices for accountability..."

Row
EC-32 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶6
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

"Addendum to GPT-5 System Card: Sensitive conversations"

Row
EC-33 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.2 ¶7
Substrate
[SAFE-HUB]
Claim type
factual
Verdict
Verified

"safety-by-design... anchored by our goal of creating a safe and engaging experience"

Row
EC-34 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.3 ¶3
Substrate
[CAI-SAFE]
Claim type
factual
Verdict
Verified

"Two and a half years later, our honest assessment..."

Row
EC-4 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶5
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"establishes the nation's first frontier AI safety and transparency requirements..."

Row
EC-5 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶6
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"how we assess and mitigate cyber offense, chemical, biological..."

Row
EC-6 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶6
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"the FCF will serve as our compliance framework for SB 53..."

Row
EC-7 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶7
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"By formalizing achievable transparency practices that responsible labs already voluntarily follow..."

Row
EC-8 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶7
Substrate
[SB53]
Claim type
factual
Verdict
Verified

"Launch 'moonshot R&D' projects..."

Row
EC-9 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.1 ¶8
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"The science of model evaluation isn't well-developed enough to provide dispositive answers"

Row
EC-1 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶3
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"approached" / "passed" RSP thresholds

Row
EC-2 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶3
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"taken a precautionary approach and implemented the relevant safeguards, but our internal uncertainty translates into a weak external case..."

Row
EC-3 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶3
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"Red teaming on its own is not a panacea for risk assessment"

Row
EC-4 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶5
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

"Risks surfaced in one point in time red teaming effort may be under-assessed..."

Row
EC-5 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶5
Substrate
[RT-PAPER]
Claim type
factual
Verdict
Verified

RAND analysis: SL5 standard "currently not possible" / "will likely require assistance from the national security community"

Row
EC-6 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶6
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"requirements at the higher RSP levels that are very hard to meet unilaterally..."

Row
EC-7 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶6
Substrate
[RSP3]
Claim type
factual
Verdict
Verified

"It should be an explicit violation of law for a lab to lie about compliance with its framework or punish employees who raise concerns about violations."

Row
EC-1 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Anthropic-federal-framework paragraph
Substrate
WAREHOUSE-PINNED [www_anthropic_com_4d535568cb]
Source
https://www.anthropic.com/news/compliance-framework-SB53
Claim type
factual (regulatory-principle articulation)
Verdict
Verified

Anthropic FCF documents the *output* of cross-functional engagement (the framework itself; the FCF compliance text) but does not document the *texture* of that engagement (consultation frequency, decision-authority allocation, disagreement-resolution process)

Row
EC-10 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Anthropic-federal-framework second paragraph
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part V §V.1 (Anthropic FCF) grounding)
Claim type
factual (substrate-character claim)
Verdict
Verified

Anthropic CEO Dario Amodei essay "Machines of Loving Grace" reference at §VII.1 Amodei-supplementary paragraph + §VIII.3.1 listing

Row
EC-17 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 + §VIII.3.1 listing
Substrate
URL substrate (darioamodei.com/essay/machines-of-loving-grace; WAREHOUSE-PINNED [www_darioamodei_com_3633040fb7])
Source
https://www.darioamodei.com/essay/machines-of-loving-grace
Claim type
factual (reference attribution; Q14-tangential supplementary substrate)
Verdict
Verified

"Yet broad confidentiality agreements block us from voicing our concerns, except to the very companies that may be failing to address these issues."

Row
EC-2 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Right-to-Warn paragraph
Substrate
WAREHOUSE-PINNED [righttowarn_ai_f1494f4b1e]
Source
https://righttowarn.ai/
Claim type
factual (structural-mechanics claim)
Verdict
Verified

"Ordinary whistleblower protections are insufficient because they focus on illegal activity, whereas many of the risks we are concerned about are not yet regulated."

Row
EC-3 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Right-to-Warn paragraph
Substrate
WAREHOUSE-PINNED [righttowarn_ai_f1494f4b1e]
Source
https://righttowarn.ai/
Claim type
factual (regulatory-coverage-gap claim)
Verdict
Verified

"Some of us reasonably fear various forms of retaliation, given the history of such cases across the industry."

Row
EC-4 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Right-to-Warn paragraph
Substrate
WAREHOUSE-PINNED [righttowarn_ai_f1494f4b1e]
Source
https://righttowarn.ai/
Claim type
factual (retaliation-history claim)
Verdict
Verified

"Right to Warn" open letter, titled "A Right to Warn about Advanced Artificial Intelligence," signed by current and former employees of frontier AI companies; signatory list visible on landing page; retrieved into research corpus 2026-05-09

Row
EC-5 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Right-to-Warn paragraph
Substrate
WAREHOUSE-PINNED [righttowarn_ai_f1494f4b1e]
Source
https://righttowarn.ai/
Claim type
factual (existence + authorship + title + retrieval-date)
Verdict
Verified

Right to Warn letter four-principles enumeration — (a) not entering or enforcing non-disparagement agreements that prohibit risk-related criticism; (b) facilitating verifiably anonymous risk-concern channel to board / regulators / independent organization; (c) supporting open-criticism culture with appropriate trade-secret protection; (d) not retaliating against public risk-related disclosure after internal processes fail

Row
EC-6 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Right-to-Warn four-principles paragraph
Substrate
WAREHOUSE-PINNED [righttowarn_ai_f1494f4b1e]
Source
https://righttowarn.ai/
Claim type
factual (claim about letter's specific demands)
Verdict
Verified

Anthropic CEO Dario Amodei essay "Machines of Loving Grace" provides current-frontier-lab-CEO perspective on AI safety + capability trajectory — supplementary lab-leadership context (Q14-tangential not Q14-substantive)

Row
EC-7 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Amodei-supplementary paragraph
Substrate
WAREHOUSE-PINNED [www_darioamodei_com_3633040fb7]
Source
https://www.darioamodei.com/essay/machines-of-loving-grace
Claim type
factual (existence + characterization)
Verdict
Verified

§VIII.2.L6 OpenAI Safety Hub "Addendum to GPT-5 System Card: Sensitive conversations" reference

Row
EC-10 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.2.L6
Substrate
WAREHOUSE-PINNED [openai_com_838d7a59a0]
Source
https://openai.com/safety/
Claim type
factual (listing reference)
Verdict
Verified

Anthropic "The need for transparency in Frontier AI" publication; citation reference at §VIII.3.1 listing

Row
EC-21 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.1 listing
Substrate
n/a (URL substrate: anthropic.com/news/the-need-for-transparency-in-frontier-ai; WAREHOUSE-PINNED [www_anthropic_com_9f010bf125])
Source
https://www.anthropic.com/news/the-need-for-transparency-in-frontier-ai
Claim type
factual (reference attribution)
Verdict
Verified

Google AI Safety hub publication; citation reference at §VIII.3.1 listing

Row
EC-22 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.1 listing
Substrate
n/a (URL substrate: ai.google/safety; WAREHOUSE-PINNED [ai_google_5ab1558b6f])
Source
https://ai.google/safety/
Claim type
factual (reference attribution)
Verdict
Verified

Volokh / Lemley / Henderson, "Freedom of Speech and AI Output," Journal of Free Speech Law (2023)

Row
EC-18 (Part I — Part I -- The Cases)
Body location
§I.7 GRAD-INTERN
Substrate
WAREHOUSE-PINNED [journaloffreespeechlaw_o_volokhlemleyhenderson_pd_ecaeba6fc1] (Branch-1 cross-warehouse: Steven_Master_Portal ai_speech_pinned; content-verified 2026-05-22)
Source
https://www.journaloffreespeechlaw.org/volokhlemleyhenderson.pdf
Claim type
factual (publication-existence) + inferential (alignment claim)
Verdict
Provisional

Peter Salib, Washington University Law Review (2024)

Row
EC-19 (Part I — Part I -- The Cases)
Body location
§I.7 GRAD-INTERN
Substrate
WAREHOUSE-PINNED [wustllawreview_org_11_Salib_FINAL-10_2_24-1_bceebb7f8e] (Branch-1 cross-warehouse: Steven_Master_Portal ai_speech_pinned; content-verified 2026-05-22) (https://wustllawreview.org/wp-content/uploads/2024/11/11_Salib_FINAL-10.2.24-1.pdf)
Source
https://wustllawreview.org/wp-content/uploads/2024/11/11_Salib_FINAL-10.2.24-1.pdf
Claim type
factual (publication-existence) + inferential (alignment claim)
Verdict
Provisional

Austin & Levy, Stanford Law Review (2025)

Row
EC-20 (Part I — Part I -- The Cases)
Body location
§I.7 GRAD-INTERN
Substrate
WAREHOUSE-PINNED [sw_1a_b1_warehouse_part1] (Branch-1 cross-warehouse: Steven_First_Amendment_Consequences; content-verified 2026-05-22 -- source_id review_law_stanford__21c3cbfe8d, Austin-Levy 77 Stan. L. Rev. 1) (https://review.law.stanford.edu/wp-content/uploads/sites/3/2025/01/Austin-Levy-77-Stan.-L.-Rev.-1.pdf)
Source
https://review.law.stanford.edu/wp-content/uploads/sites/3/2025/01/Austin-Levy-77-Stan.-L.-Rev.-1.pdf
Claim type
factual (publication-existence) + inferential (alignment claim)
Verdict
Provisional

Bommasani et al. 2021 *On the Opportunities and Risks of Foundation Models* (Stanford CRFM) introduces the term "foundation model"

Row
EC-1 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 2
Substrate
chunk-hash `a9b3bcee`
Source
https://arxiv.org/pdf/2108.07258.pdf
Claim type
factual (publication-existence + concept-introduction)
Verdict
Verified

"262 assets, 356 dependencies in early 2023" cited again at §II.2 GRAD-INTERN as evidence for granularity claim

Row
EC-17 (Part II — Part II -- The Ecosystem)
Body location
§II.2 GRAD-INTERN
Substrate
chunk-hash `d0ef4c36` (re-cited per EC-7)
Claim type
factual (quantitative-figures referenced for inferential warrant in granularity argument)
Verdict
Verified

"The defects of the foundation model are inherited by all the adapted models downstream" — Bommasani 2021 verbatim quote

Row
EC-2 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 2
Substrate
chunk-hash `a9b3bcee`
Claim type
factual (verbatim-quoted concept)
Verdict
Verified

"BERT, GPT-3, DALL-E" as foundation-model examples

Row
EC-3 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 2
Substrate
chunk-hash `a9b3bcee`
Claim type
factual (model-enumeration; common knowledge in ML field)
Verdict
Verified

Bommasani et al. 2023 *Ecosystem Graphs: The Social Footprint of Foundation Models* (Stanford CRFM)

Row
EC-4 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 3
Substrate
chunk-hash `d0ef4c36`
Source
https://arxiv.org/abs/2303.15772
Claim type
factual (publication-existence)
Verdict
Verified

"broader sociotechnical ecosystem" — Bommasani et al. 2023 verbatim quote

Row
EC-5 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 3
Substrate
chunk-hash `d0ef4c36`
Claim type
factual (verbatim-quoted concept)
Verdict
Verified

"assets (datasets, models, applications) linked together by dependencies that indicate technical (e.g. how Bing relies on GPT-4) and social (e.g. how Microsoft relies on OpenAI) relationships" — Bommasani et al. 2023 verbatim quote with embedded examples

Row
EC-6 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 3
Substrate
chunk-hash `d0ef4c36`
Claim type
factual (verbatim-quoted methodology + factual dependency examples)
Verdict
Verified

"As of March 2023 the paper documents 262 assets (64 datasets, 128 models, 70 applications) from 63 organizations linked by 356 dependencies" — Bommasani et al. 2023 figures

Row
EC-7 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 3
Substrate
chunk-hash `d0ef4c36`
Claim type
factual (quantitative figures)
Verdict
Verified

Stanford Center for Research on Foundation Models (CRFM) authorship affiliation

Row
EC-8 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraphs 2-3
Substrate
chunk-hashes `a9b3bcee` + `d0ef4c36`
Claim type
factual (institutional affiliation)
Verdict
Verified

Austin & Levy Stanford Law Review (2025)

Row
EC-10 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.b
Substrate
WAREHOUSE-PINNED [sw_1a_b1_warehouse_part1] (https://review.law.stanford.edu/wp-content/uploads/sites/3/2025/01/Austin-Levy-77-Stan.-L.-Rev.-1.pdf)
Source
https://review.law.stanford.edu/wp-content/uploads/sites/3/2025/01/Austin-Levy-77-Stan.-L.-Rev.-1.pdf
Claim type
factual (publication) + inferential
Verdict
Provisional

Harvard Law Review "Beyond §230 Principles for AI Governance" vol 138 (2025)

Row
EC-11 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.2 + §IV.3.b + §IV.3.c
Substrate
WAREHOUSE-PINNED [sw_1a_b1_warehouse_part1] (https://harvardlawreview.org/print/vol-138/beyond-section-230-principles-for-ai-governance/)
Source
https://harvardlawreview.org/print/vol-138/beyond-section-230-principles-for-ai-governance/
Claim type
factual (publication) + inferential
Verdict
Provisional

Cardozo Law Review deepfakes-1A article

Row
EC-12 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.b
Substrate
WAREHOUSE-PINNED [sw_1a_b1_warehouse_part1] (https://cardozolawreview.com/deeply-fake-deeply-disturbing-deeply-constitutional-why-the-first-amendment-likely-protects-the-creation-of-pornographic-deepfakes/)
Source
https://cardozolawreview.com/deeply-fake-deeply-disturbing-deeply-constitutional-why-the-first-amendment-likely-protects-the-creation-of-pornographic-deepfakes/
Claim type
factual (publication)
Verdict
Provisional

Harvard JOLT generative-AI-§230 article

Row
EC-13 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.b + §IV.3.c
Substrate
WAREHOUSE-PINNED [sw_1a_b1_warehouse_part1] (https://jolt.law.harvard.edu/digest/generative-ai-will-break-the-internet-beyond-section-230)
Source
https://jolt.law.harvard.edu/digest/generative-ai-will-break-the-internet-beyond-section-230
Claim type
factual (publication)
Verdict
Provisional

§IV.3.c §230 academic-literature multi-source list (CDT analytical piece + Moody's §230 analysis + Journal of Free Speech Law "Where's the Liability in Harmful AI Speech?") — disaggregation

Row
EC-14 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.c
Substrate
WAREHOUSE-PINNED [multi-source]: (a) CDT "Section 230 and Its Applicability to Generative AI: A Legal Analysis" pinned at Anthropic Counsel Research WarehouseA; (b) Moody's [pin-12144a1a] already pinned; (c) Journal of Free Speech Law "Where's the Liability?" Henderson/Hashimoto/Lemley 2023 pinned at AIH aih_p5_7_b0_pdfs_r1_warehouse_part1.json
Source
https://www.moodys.com/web/en/us/insights/insurance/230-immunity-for-AI-chatbot-lawsuits.html   https://cdt.org/insights/section-230-and-its-applicability-to-generative-ai-a-legal-analysis/
Claim type
factual (publication-existence) + inferential (§230-doctrinal-misfit alignment)
Verdict
Verified

Lidsky and Daves (Journal of Free Speech Law)

Row
EC-26 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.b academic-line list
Substrate
WAREHOUSE-PINNED [sw_1a_b1_warehouse_part1] (https://www.journaloffreespeechlaw.org/lidskydaves.pdf)
Source
https://www.journaloffreespeechlaw.org/lidskydaves.pdf
Claim type
factual (publication)
Verdict
Provisional

Volokh-Lemley-Henderson "Freedom of Speech and AI Output" Journal of Free Speech Law (2023)

Row
EC-8 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.b
Substrate
WAREHOUSE-PINNED [journaloffreespeechlaw_o_volokhlemleyhenderson_pd_ecaeba6fc1] (https://www.journaloffreespeechlaw.org/volokhlemleyhenderson.pdf)
Source
https://www.journaloffreespeechlaw.org/volokhlemleyhenderson.pdf
Claim type
factual (publication) + inferential (alignment claim)
Verdict
Provisional

Peter Salib WULR (2024)

Row
EC-9 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §IV.3.b
Substrate
WAREHOUSE-PINNED [wustllawreview_org_11_Salib_FINAL-10_2_24-1_bceebb7f8e] (https://wustllawreview.org/wp-content/uploads/2024/11/11_Salib_FINAL-10.2.24-1.pdf)
Source
https://wustllawreview.org/wp-content/uploads/2024/11/11_Salib_FINAL-10.2.24-1.pdf
Claim type
factual (publication) + inferential (alignment)
Verdict
Provisional

NCSL AI Legislation Tracker (2024; STALE at 2026-05-06 access date)

Row
EC-32 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.4.f
Substrate
URL-only (https://www.ncsl.org/technology-and-communication/artificial-intelligence-2024-legislation)
Source
https://www.ncsl.org/technology-and-communication/artificial-intelligence-2024-legislation
Claim type
factual (publication-existence + staleness-acknowledgment)
Verdict
Verified

EU AI Act tracker (artificialintelligenceact.eu) reference at §IV.1 federal-preemption discussion + §VIII.3.1 listing

Row
EC-39 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.1 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [artificialintelligenceac_1ad6dc89cd]
Source
https://artificialintelligenceact.eu/
Claim type
factual (reference attribution)
Verdict
Verified

OpenAI denial: cited pre-existing suicidal ideation; framed user as "tricking" the model via fictional-character framing

Row
EC-10 (Part I — Part I -- The Cases)
Body location
§I.3 paragraph 2
Substrate
WAREHOUSE-PINNED [www_nbcnews_com_b795cee9f3]
Source
https://www.nbcnews.com/tech/tech-news/openai-denies-allegation-chatgpt-teenagers-death-adam-raine-lawsuit-rcna245946
Claim type
factual (defendant-position-as-stated)
Verdict
FLAG-B-K (per EC-4)

Amended-complaint allegation: "OpenAI removed safeguards" before Adam's death

Row
EC-12 (Part I — Part I -- The Cases)
Body location
§I.3 paragraph 4
Substrate
WAREHOUSE-PINNED [time_com_9ff8b444d3]
Source
https://time.com/7327946/chatgpt-openai-suicide-adam-raine-lawsuit/
Claim type
factual (allegation as alleged)
Verdict
Verified

November 2025 college-graduate suit against OpenAI; case caption + court + docket UNSURFACED at access date

Row
EC-14 (Part I — Part I -- The Cases)
Body location
§I.4 entire section
Substrate
WAREHOUSE-PINNED [www_cnn_com_5ba50be091]
Source
https://www.cnn.com/2025/11/06/us/openai-chatgpt-suicide-lawsuit-invs-vis
Claim type
factual (procedural)
Verdict
Verified

PA AG sued Character Tech; complaint alleges fabricated psychiatric credentials

Row
EC-15 (Part I — Part I -- The Cases)
Body location
§I.5 paragraph 1
Substrate
WAREHOUSE-PINNED [multi-source: thenextweb_com_a022a2fb4e + www_attorneygeneral_gov_9783df38e4]
Source
https://thenextweb.com/news/pennsylvania-character-ai-chatbot-doctor-lawsuit   https://www.attorneygeneral.gov/taking-action/ag-henry-files-lawsuit-against-character-ai/
Claim type
factual (procedural) + factual (substantive allegation)
Verdict
Verified

Four settled NY/CO/TX matters; individual captions/dockets/decedent identifications UNSURFACED

Row
EC-16 (Part I — Part I -- The Cases)
Body location
§I.6 paragraph 1
Substrate
WAREHOUSE-PINNED [multi-source: www_cnn_com_03ed409528 + www_cnbc_com_a0b5e52ad2]
Source
https://www.cnn.com/2026/01/07/business/character-ai-google-settle-teen-suicide-lawsuit   https://www.cnbc.com/2026/01/07/google-characterai-to-settle-suits-involving-suicides-ai-chatbots.html
Claim type
factual (procedural)
Verdict
Verified

NBC News coverage of the Raine family wrongful-death lawsuit against OpenAI; citation reference at §I.3 + §VIII.3.1 listing

Row
EC-29 (Part I — Part I -- The Cases)
Body location
§I.3 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [www_nbcnews_com_5103552a88]
Source
https://www.nbcnews.com/tech/tech-news/family-teenager-died-suicide-alleges-openais-chatgpt-blame-rcna226147
Claim type
factual (reference attribution)
Verdict
Verified

TheNextWeb coverage of Pennsylvania AG action against Character.AI for unauthorized practice of medicine; citation reference at §I.5 + §VIII.3.1 listing

Row
EC-32 (Part I — Part I -- The Cases)
Body location
§I.5 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [thenextweb_com_a022a2fb4e]
Source
https://thenextweb.com/news/pennsylvania-character-ai-chatbot-doctor-lawsuit
Claim type
factual (reference attribution)
Verdict
Verified

CNN coverage (2025-08-26) of OpenAI/ChatGPT teen suicide lawsuit filing; citation reference at §I.3 + §VIII.3.1 listing

Row
EC-33 (Part I — Part I -- The Cases)
Body location
§I.3 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [www_cnn_com_388418d213]
Source
https://www.cnn.com/2025/08/26/tech/openai-chatgpt-teen-suicide-lawsuit
Claim type
factual (reference attribution)
Verdict
Verified

CNN coverage (2025-11-06) of OpenAI/ChatGPT suicide case (Nov 2025 college-graduate case); citation reference at §I.4 + §VIII.3.1 listing

Row
EC-34 (Part I — Part I -- The Cases)
Body location
§I.4 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [www_cnn_com_5ba50be091]
Source
https://www.cnn.com/2025/11/06/us/openai-chatgpt-suicide-lawsuit-invs-vis
Claim type
factual (reference attribution)
Verdict
Verified

Garcia + 4 additional matters settled January 2026 (NY / CO / TX; CO 13-year-old Thornton)

Row
EC-6 (Part I — Part I -- The Cases)
Body location
§I.2 paragraph 4 + §I.6
Substrate
WAREHOUSE-PINNED [www_cnn_com_03ed409528 + www_cnbc_com_a0b5e52ad2 + www_jurist_org_5c7eb8734a + www_claimsjournal_com_e75482a53f]
Source
https://www.cnn.com/2026/01/07/business/character-ai-google-settle-teen-suicide-lawsuit   https://www.jurist.org/news/2026/01/google-and-character-ai-agree-to-settle-lawsuit-linked-to-teen-suicide/   https://www.cnbc.com/2026/01/07/google-characterai-to-settle-suits-involving-suicides-ai-chatbots.html   https://www.claimsjournal.com/news/national/2026/01/08/334964.htm
Claim type
factual (procedural)
Verdict
Verified

Raine amended-complaint allegation "OpenAI removed safeguards before the decedent's death" cross-reference applied to Layer 2 framing

Row
EC-7 (Part III-a — Part III -- Layers 1-3)
Body location
§III.2 legal-issues paragraph re post-sale-modification doctrinal frontier
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I EC-12 (Raine amended-complaint) grounding)
Claim type
factual (allegation as cross-referenced)
Verdict
Verified

Jurist: Google + Character.AI settled Garcia 2026-01-08; component-part-manufacturer theory + "*strict liability... 'for failing to prevent harm to minors arising from their foreseeable use of such products.'*"

Row
EC-11 (Part III-b — Part III -- Layers 4-7)
Body location
§III.5 ¶2 + §III.5 ¶3 bullet 1
Substrate
WAREHOUSE-PINNED [www_jurist_org_5c7eb8734a]
Source
https://www.jurist.org/news/2026/01/google-and-character-ai-agree-to-settle-lawsuit-linked-to-teen-suicide/
Claim type
factual (procedural + allegation)
Verdict
Verified

Psychiatric Times: OpenAI disclosure "*approximately 1.2 million of its 800 million ChatGPT users discuss suicide weekly on its platform*" (citing OpenAI 2025-10)

Row
EC-13 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶2
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (lab-disclosed statistic)
Verdict
Verified

Psychiatric Times: "*Character.AI responded by banning minors from using open-ended chats with their product in November 2025*" (citing AP 2025-10-29)

Row
EC-14 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶3 bullet 2
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (deployer product change)
Verdict
Verified

Psychiatric Times: AI sycophancy "*the tendency towards indiscriminately reinforcing a user's ideas*" + OCD + eating-disorders implications

Row
EC-15 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶3 bullet 5 + §III.6 GRAD-INTERN
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (clinical-pattern observation)
Verdict
Provisional

Psychiatric Times: Conway ruling "*In the initial phase of the trial, a federal judge ruled against Character Technologies' argument that its chatbot's output was protected by the First Amendment or that it constituted 'speech' at all.*"

Row
EC-17 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶4 bullet 3
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (judicial-holding via secondary cite)
Verdict
Verified

Psychiatric Times re *Commonwealth v Carter* (2017): "*That case set the new precedent that words are not always protected by the First Amendment and could effectively be treated as a murder weapon.*"

Row
EC-18 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶4 bullet 4
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (commentary precedent characterization)
Verdict
Verified

Psychiatric Times: KY AG action against Character Tech "*harmful, explicit, and psychologically manipulative interactions with minors*" (citing Kentucky.gov 2026-01-08)

Row
EC-19 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶4 bullet 3
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (allegation as alleged)
Verdict
Verified

Psychiatric Times re Meta+Google LA social-media-addiction verdict (2026-03; $6M) + Meta NM AG action (2026-03; $375M penalty)

Row
EC-20 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶4 bullet 5
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (verdict-existence + amount)
Verdict
Verified

Psychiatric Times re future-litigation theory: "*chatbot use caused or exacerbated other mental disorders such as obsessive-compulsive disorder or eating disorders, given the tendency towards indiscriminately reinforcing a user's ideas (AI sycophancy)*"

Row
EC-21 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 GRAD-INTERN ¶1
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (commentary forecast of future-litigation theory)
Verdict
FLAG-B-K (per EC-15)

Psychiatric Times re forensic-psychiatry expert role: evaluate "*symptoms, reach diagnoses, opine about damages, and potentially attempt to evaluate causation*" + "*in the absence of robust research... the [causation function] may prove more challenging*"

Row
EC-22 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 GRAD-INTERN ¶1 closing
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (commentary on expert-role landscape)
Verdict
Verified

Q14 cross-functional coordination question UNANSWERED at R1 + Psychiatric Times: "*AI companies are spending fortunes in a successful effort to avoid government regulation*" (oblique evidence)

Row
EC-29 (Part III-b — Part III -- Layers 4-7)
Body location
§III.7 ¶3 bullet 5
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (commentary observation) + scope-statement (R1 unanswered)
Verdict
Verified

Character.AI January 2026 settlement of Garcia + multiple matters + ongoing model roadmap activity through 2026-04

Row
EC-9 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 GRAD-INTERN ¶1
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I EC-6 + §I.2 / §I.6 grounding)
Claim type
factual (procedural cross-reference)
Verdict
Verified

Raine amended-complaint safeguard-removal allegation

Row
EC-28 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.2 + §IV.3.a + §IV.3.e
Substrate
WAREHOUSE-PINNED [time_com_9ff8b444d3]
Source
https://time.com/7327946/chatgpt-openai-suicide-adam-raine-lawsuit/
Claim type
factual (allegation as alleged)
Verdict
Verified

PA AG matter (consumer-protection + unauthorized-practice-of-medicine framing)

Row
EC-35 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.f
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I §I.5 grounding)
Claim type
factual (procedural + substantive allegation)
Verdict
Verified

SB3444 "would block certain civil lawsuits against a 'developer' of a frontier model for specific types of 'critical harms'..."

Row
EC-10 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶3
Substrate
[HOOD]
Claim type
factual
Verdict
Verified

"Critical harm" defined as "incidents involving the death or serious injury of 100 or more people, or at least $1,000,000,000 in property damage"

Row
EC-11 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶3
Substrate
[HOOD]
Claim type
definitional
Verdict
Verified

SB3261 "would put more teeth into the rules. Frontier developers and large chatbot providers would be required to publish public safety..."

Row
EC-12 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶3
Substrate
[HOOD]
Claim type
factual
Verdict
Verified

Anthropic statement: "We are opposed to this bill" / "transparency laws should pair public safety with genuine accountability..."

Row
EC-13 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶4
Substrate
[HOOD]
Claim type
factual
Verdict
Verified

Woodside (Secure AI Project) via WIRED via Hoodline: "would take the extreme step of nearly eliminating liability for severe harms"

Row
EC-14 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶4
Substrate
[HOOD]
Claim type
factual
Verdict
Verified

SB3444 federal-preemption hook: "would stop applying if Congress or a federal agency adopts overlapping rules..."

Row
EC-15 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶5
Substrate
[HOOD]
Claim type
factual
Verdict
Verified

"Illinois lawmakers are weighing dueling Senate plans..."

Row
EC-9 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 ¶2
Substrate
[HOOD]
Claim type
factual
Verdict
Verified

Illinois SB 3444 / SB 3261 contest cross-reference — two major frontier labs publicly take opposing positions on a state liability-shield bill (OpenAI supporting; Anthropic opposing); frequency / structure / decision-making process behind that engagement not in public record

Row
EC-11 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Illinois-contest paragraph
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part VI §VI.2 grounding)
Claim type
factual (legislative-engagement observation)
Verdict
Verified

"AI companies are spending fortunes in a successful effort to avoid government regulation."

Row
EC-8 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.1 Psychiatric-Times paragraph
Substrate
WAREHOUSE-PINNED [www_psychiatrictimes_com_2290cfefb3]
Source
https://www.psychiatrictimes.com/view/the-psychiatrist-s-preview-of-legal-cases-against-big-ai
Claim type
factual (industry-level observation)
Verdict
Verified

insights.marvin-42.com Anthropic-context coverage of frontier-safety-roadmap; citation reference at §VIII.3.1 listing

Row
EC-20 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.1 listing
Substrate
n/a (URL substrate: insights.marvin-42.com/articles/anthropic-publishes-frontier-safety-roadmap-with-2026-2027-targets; WAREHOUSE-PINNED [insights_marvin_42_com_4eeb206c20])
Source
https://insights.marvin-42.com/articles/anthropic-publishes-frontier-safety-roadmap-with-2026-2027-targets
Claim type
factual (reference attribution)
Verdict
Verified

§VIII.1.3 Nov-2025 college-graduate case (caption/court/docket UNSURFACED)

Row
EC-4 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.1.3
Substrate
WAREHOUSE-PINNED [www_cnn_com_5ba50be091]
Source
https://www.cnn.com/2025/11/06/us/openai-chatgpt-suicide-lawsuit-invs-vis
Claim type
factual (existence + reported allegations)
Verdict
Verified

§VIII.1.4 PA AG case (full caption pending)

Row
EC-5 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.1.4
Substrate
WAREHOUSE-PINNED [thenextweb_com_a022a2fb4e]
Source
https://thenextweb.com/news/pennsylvania-character-ai-chatbot-doctor-lawsuit
Claim type
factual (existence + allegations)
Verdict
Verified

§VIII.1.5 4 settled NY/CO/TX cases (individual captions pending)

Row
EC-6 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.1.5
Substrate
WAREHOUSE-PINNED [multi-source: www_cnn_com_03ed409528 + www_cnbc_com_a0b5e52ad2]
Source
https://www.cnn.com/2026/01/07/business/character-ai-google-settle-teen-suicide-lawsuit   https://www.cnbc.com/2026/01/07/google-characterai-to-settle-suits-involving-suicides-ai-chatbots.html
Claim type
factual (existence + settlement timing)
Verdict
Verified

"By the end of 2025, at least ten families had filed lawsuits... seven... a minor at the time of their death"

Row
EC-1 (Part I — Part I -- The Cases)
Body location
§I.1 paragraph 1
Substrate
WAREHOUSE-PINNED [lawstreetmedia_com_d6e4ee7581]
Source
https://lawstreetmedia.com/insights/a-new-wave-of-litigation-over-ai-chatbots/
Claim type
factual (aggregate)
Verdict
Verified

Garcia filing date "October 2024" + Middle District of Florida + docket 6:24-cv-1903-ACC-UAM

Row
EC-2 (Part I — Part I -- The Cases)
Body location
§I.2 caption + body
Substrate
WAREHOUSE-PINNED [socialmediavictims_org_bba5c4d37c + incidentdatabase_ai_31fae1fdf0]
Source
https://socialmediavictims.org/character-ai-lawsuits/   https://incidentdatabase.ai/cite/826/
Claim type
factual (procedural)
Verdict
Verified

Component-part-manufacturer theory naming Google (Garcia)

Row
EC-22 (Part I — Part I -- The Cases)
Body location
§I.2 paragraph 2
Substrate
WAREHOUSE-PINNED [socialmediavictims_org_bba5c4d37c]
Source
https://socialmediavictims.org/character-ai-lawsuits/
Claim type
factual (allegation as alleged) + definitional (doctrinal-category invocation)
Verdict
Verified

"Roughly two-thirds majority of decedents in the broader wave are minors"

Row
EC-23 (Part I — Part I -- The Cases)
Body location
§I.6 GRAD-INTERN
Substrate
WAREHOUSE-PINNED (cross-reference: inherits Part I §I.1 / EC-1 aggregator grounding)
Claim type
factual (aggregate)
Verdict
Verified

"her son, Sewell Setzer III, who died in February 2024 at age 14 in Florida"

Row
EC-3 (Part I — Part I -- The Cases)
Body location
§I.2 paragraph 1
Substrate
WAREHOUSE-PINNED [socialmediavictims_org_bba5c4d37c]
Source
https://socialmediavictims.org/character-ai-lawsuits/
Claim type
factual (procedural)
Verdict
Verified

Transparency Coalition Barcott analysis of Conway "product-not-speech" ruling in Garcia v. Character Technologies; citation reference at §I.2 + cross-reference at §VI.3.1 + §VIII.3.1 listing

Row
EC-30 (Part I — Part I -- The Cases)
Body location
§I.2 + §VIII.3.1 listing
Substrate
WAREHOUSE-PINNED [www_transparencycoalitio_870e7ad0c9]
Source
https://www.transparencycoalition.ai/news/important-early-ruling-in-characterai-case-this-chatbot-is-a-product-not-speech
Claim type
factual (reference attribution)
Verdict
Verified

"months-long relationship with Character.AI personas... emotionally intense and... sexually explicit role-play"

Row
EC-4 (Part I — Part I -- The Cases)
Body location
§I.2 paragraph 1
Substrate
WAREHOUSE-PINNED [socialmediavictims_org_bba5c4d37c]
Source
https://socialmediavictims.org/character-ai-lawsuits/
Claim type
factual (allegation as alleged)
Verdict
FLAG-B-K (overlay narrative in body: "The suits do not allege that the chatbot held a gun" passage at §I.1 paragraph 2 partially functions as B-K overlay — distinguishes plaintiff allegation pattern from chatbot-as-meaning-engaging)

Barcott: "Significantly, Judge Conway ruled that Character.AI is a product for the purposes of product liability claims, and not a service"

Row
EC-17 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.3.1 ¶1
Substrate
[TC-BARCOTT]
Claim type
factual
Verdict
Verified

Barcott: "That doesn't necessarily mean Judge Conway decided that the Character.AI interactions with Setzer were or were not protected speech..."

Row
EC-18 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.3.1 ¶1
Substrate
[TC-BARCOTT]
Claim type
factual
Verdict
Verified

TJLP via Barcott: "sends a clear signal to companies developing and deploying LLM-powered products at scale that they cannot evade legal consequences..."

Row
EC-19 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.3.1 ¶1
Substrate
[TC-BARCOTT]
Claim type
factual
Verdict
Verified

§VIII.1.6 aggregate-meta finding (10 / 7 minors / class enumeration)

Row
EC-7 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.1.6
Substrate
WAREHOUSE-PINNED [lawstreetmedia_com_d6e4ee7581]
Source
https://lawstreetmedia.com/insights/a-new-wave-of-litigation-over-ai-chatbots/
Claim type
factual (aggregate)
Verdict
Verified

KL Gates: "*the first bellwether jury verdicts will establish critical precedents for the cases that follow.*"

Row
EC-10 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 GRAD-INTERN ¶1 closing
Substrate
WAREHOUSE-PINNED [www_klgates_com_d785c9e47c]
Source
https://www.klgates.com/AI-Product-Liability-The-Next-Wave-of-Litigation-3-27-2026
Claim type
factual (commentary forecast)
Verdict
Verified

McGuireWoods: MDL 3047 ruling "*limited Section 230's protective reach to third-party content claims, distinguishing them from claims targeting a platform's own design architecture.*"

Row
EC-16 (Part III-b — Part III -- Layers 4-7)
Body location
§III.6 ¶4 bullet 1
Substrate
WAREHOUSE-PINNED [www_mcguirewoods_com_d13082dedf]
Source
https://www.mcguirewoods.com/client-resources/alerts/2026/3/can-social-media-or-ai-be-a-defective-product/
Claim type
factual (commentary doctrinal observation)
Verdict
Verified

Conway in *Garcia v Character Technologies*: chatbot is "*a 'product for the purposes of Plaintiff's claims [arising] from defects in the Character A.I. app rather than ideas or expressions within the app.'*"

Row
EC-2 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 1
Substrate
WAREHOUSE-PINNED [www_mcguirewoods_com_d13082dedf]
Source
https://www.mcguirewoods.com/client-resources/alerts/2026/3/can-social-media-or-ai-be-a-defective-product/
Claim type
factual (judicial holding)
Verdict
Verified

McGuireWoods: "*the learned intermediary doctrine offers no refuge for direct-to-consumer platforms marketed to and used by minors*"

Row
EC-3 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 2
Substrate
WAREHOUSE-PINNED [www_mcguirewoods_com_d13082dedf]
Source
https://www.mcguirewoods.com/client-resources/alerts/2026/3/can-social-media-or-ai-be-a-defective-product/
Claim type
factual (commentary doctrinal observation)
Verdict
Verified

KL Gates: "*Plaintiffs... increasingly draft complaints to target the architecture of the deployed system — guardrails, defaults, escalation pathways, and marketing — so the case looks like a product-defect dispute instead of a content dispute.*"

Row
EC-4 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 3
Substrate
WAREHOUSE-PINNED [www_klgates_com_d785c9e47c]
Source
https://www.klgates.com/AI-Product-Liability-The-Next-Wave-of-Litigation-3-27-2026
Claim type
factual (commentary doctrinal observation)
Verdict
Verified

KL Gates: "*liability theories are moving up and down the AI supply chain as plaintiffs explore component-part and substantial-participation theories that can reach upstream and downstream actors.*"

Row
EC-5 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 4 + §III.5 ¶3 bullet 2
Substrate
WAREHOUSE-PINNED [www_klgates_com_d785c9e47c]
Source
https://www.klgates.com/AI-Product-Liability-The-Next-Wave-of-Litigation-3-27-2026
Claim type
factual (commentary observation)
Verdict
Verified

McGuireWoods: "*the continuous-update nature of software products creates plausible grounds for courts to impose an ongoing post-sale duty to warn and implement safety features as evidence of harm accumulates.*"

Row
EC-7 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 6
Substrate
WAREHOUSE-PINNED [www_mcguirewoods_com_d13082dedf]
Source
https://www.mcguirewoods.com/client-resources/alerts/2026/3/can-social-media-or-ai-be-a-defective-product/
Claim type
factual (commentary doctrinal observation)
Verdict
Verified

MDL No. 3047 *In re Social Media Adolescent Addiction* — 2,200+ active cases as of 2026-02; N.D. Cal.; bellwether-rulings-as-persuasive-authority framing

Row
EC-8 (Part III-b — Part III -- Layers 4-7)
Body location
§III.4 ¶6 bullet 7
Substrate
WAREHOUSE-PINNED [www_mcguirewoods_com_d13082dedf + www_klgates_com_d785c9e47c]
Source
https://www.klgates.com/AI-Product-Liability-The-Next-Wave-of-Litigation-3-27-2026   https://www.mcguirewoods.com/client-resources/alerts/2026/3/can-social-media-or-ai-be-a-defective-product/
Claim type
factual (procedural-landscape)
Verdict
Verified

Tyson Mendes Raine matter commentary

Row
EC-25 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.a
Substrate
WAREHOUSE-PINNED [tysonmendes_com_73d5059ce9]
Source
https://www.tysonmendes.com/raine-v-openai-ai-product-liability-lawsuit/
Claim type
factual (publication)
Verdict
Verified

CCIA "Preemption Report 2025"

Row
EC-31 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.2 + §IV.3.d
Substrate
superseded — §IV.3.d use replaced by EC-38 (CCIA "$600 Billion Abundance Dividend" article [pin-d84eb5bc]); §IV.2 use carried no live body citation
Source
https://ccianet.org/articles/600-billion-ai-abundance-dividend-from-federal-preemption-of-state-laws/
Claim type
factual (publication) + inferential (industry-favored position)
Verdict
Provisional

§IV.4.a Goodwin Procter "EU Updates its Product Liability Regime: Important Considerations for Providers of AI Systems and Software" (Feb 18 2025) — Brussels-effect / regulatory-floor substrate

Row
EC-37 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.4.a paragraph 2
Substrate
WAREHOUSE-PINNED [pin-4cbde1b2]
Source
https://www.goodwinlaw.com/en/insights/publications/2025/02/alerts-practices-aiml-eu-updates-its-product-liability-regime
Claim type
factual (commentary citation) + inferential (regulatory-floor claim grounded against silver substrate)
Verdict
Provisional

CCIA "$600 Billion AI Abundance Dividend from Federal Preemption of State Laws" (Trevor Wagener, Nov 28, 2025)

Row
EC-38 (Part IV — Part IV -- The Legal Landscape)
Body location
§IV.3.d
Substrate
warehouse-pinned `aih_p5_7_b1_ccia_r1` [pin-d84eb5bc] (88 artifacts; 17,670 char extracted; 135,202 byte source)
Claim type
factual (publication) + inferential (industry-favored position on the preemption question)
Verdict
Verified

"the first bellwether jury verdicts will establish critical precedents for the cases that follow"

Row
EC-35 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.3 GRAD-INTERN
Substrate
[KLG]
Claim type
factual
Verdict
Verified

Conway holding: "a 'product for the purposes of Plaintiff's claims [arising] from defects in the Character A.I. app rather than ideas or expressions within the app'"

Row
EC-16 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.3.1 ¶1
Substrate
[MGW]
Claim type
factual
Verdict
Verified

K&L Gates: "liability theories are moving up and down the AI supply chain as plaintiffs explore component-part and substantial-participation theories..."

Row
EC-20 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.3.2 ¶1
Substrate
[KLG]
Claim type
factual
Verdict
Verified

K&L Gates: "the first bellwether jury verdicts will establish critical precedents for the cases that follow"

Row
EC-21 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.2 GRAD-INTERN + §VI.5 ¶2
Substrate
[KLG]
Claim type
factual
Verdict
Verified

K&L Gates: "Plaintiffs, by contrast, increasingly draft complaints to target the architecture of the deployed system—guardrails, defaults, escalation pathways, and marketing—so the case looks like a product-defect dispute instead of a content dispute"

Row
EC-22 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.3 GRAD-INTERN
Substrate
[KLG]
Claim type
factual
Verdict
Verified

"the continuous-update nature of software products creates plausible grounds for courts to impose an ongoing post-sale duty to warn..."

Row
EC-8 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.1 ¶7
Substrate
[MGW]
Claim type
factual
Verdict
Verified

§VIII.2.L5 Conway opinion citation form "785 F. Supp. 3d 1157, 1180 (M.D. Fla. 2025) per McGuireWoods coverage; primary court-document PDF not directly retrieved in research corpus"

Row
EC-9 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.2.L5
Substrate
WAREHOUSE-PINNED [www_mcguirewoods_com_d13082dedf]
Source
https://www.mcguirewoods.com/client-resources/alerts/2026/3/can-social-media-or-ai-be-a-defective-product/
Claim type
factual (citation form)
Verdict
Verified

WaPo profile late 2025 (paywalled at access date; not used as substrate)

Row
EC-13 (Part I — Part I -- The Cases)
Body location
§I.3 paragraph 5
Substrate
explicit-flag (paywalled; acknowledged, not used as substrate)
Claim type
factual (non-citation note)
Verdict
Verified

Legal-theories enumeration (wrongful death / product liability / negligence / failure to warn / sexual abuse-grooming / IIED / consumer-protection)

Row
EC-21 (Part I — Part I -- The Cases)
Body location
§I.1 paragraph 3
Substrate
inferential (synthesis grounded in cited Garcia + Raine complaints + PA AG action)
Claim type
inferential (synthesis-of-pleaded-theories)
Verdict
Verified

"the Eleventh" (federal trial court within Eleventh Circuit; appellate review not yet decided)

Row
EC-25 (Part I — Part I -- The Cases)
Body location
§I.7 paragraph 4
Substrate
inferential (general legal knowledge — federal circuit assignment)
Claim type
factual (definitional)
Verdict
Verified

Class A/B/C/D taxonomy of case-class structure

Row
EC-26 (Part I — Part I -- The Cases)
Body location
§I.1 GRAD-INTERN
Substrate
inferential (categorical synthesis)
Claim type
definitional (analytical taxonomy)
Verdict
Verified

"We do not paraphrase from the WaPo piece directly"

Row
EC-27 (Part I — Part I -- The Cases)
Body location
§I.3 paragraph 5
Substrate
explicit-flag (explicit non-citation disclosure)
Claim type
scope-statement
Verdict
Verified

Seven-layer reshape as Claude-authored analytical contribution

Row
EC-10 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraph 4 + §II.3 opening + §II.3 GRAD-INTERN
Substrate
inferential (Claude-authored analytical contribution; explicitly framed as 'our seven-layer reshape')
Claim type
definitional (analytical taxonomy)
Verdict
Verified

RLHF, Constitutional AI, RLAIF as alignment-training methods

Row
EC-11 (Part II — Part II -- The Ecosystem)
Body location
§II.2 bullet "Alignment training methodology" + §II.3 Layer 2
Substrate
inferential (general field-knowledge)
Claim type
factual (method-enumeration)
Verdict
Verified

Part II §II.1 "stack of layers, each operated by different teams (often at different companies, often in different jurisdictions)" — multi-actor multi-jurisdiction characterization

Row
EC-18 (Part II — Part II -- The Ecosystem)
Body location
§II.1 paragraph 2
Substrate
inferential (synthesis grounded in Bommasani et al. -- Part II EC-6 / EC-7 warehouse substrate)
Claim type
factual (structural-characterization of ecosystem)
Verdict
Provisional

"Pre-training-data curation choices about whether to filter suicide-related text from the corpus, whether to include content from clinical contexts, whether to distinguish first-person ideation from third-person discussion" — pre-training data design choices enumeration

Row
EC-19 (Part II — Part II -- The Ecosystem)
Body location
§II.3 Layer 1 description
Substrate
inferential (synthesis; pre-training-corpus design choices)
Claim type
factual (design-choice enumeration)
Verdict
Verified

Bonus question Q14 "department-counsel coordination gaps" cross-reference at Layer 7

Row
EC-22 (Part II — Part II -- The Ecosystem)
Body location
§II.3 Layer 7 description
Substrate
inferential (project-internal cross-reference to parent AIH MMP §J Q14; not an external source)
Claim type
factual (research-question-existence)
Verdict
Verified

"academic ecosystem-analysis tradition" framing (analytical-mode characterization)

Row
EC-9 (Part II — Part II -- The Ecosystem)
Body location
§II.2 paragraphs 1-4
Substrate
inferential (synthesis grounded in EC-1 + EC-4 substrate)
Claim type
factual (mode-existence)
Verdict
Verified

Pre-training data curation as "causally upstream of every output" with corpus-content-affecting-downstream-behavior framing

Row
EC-1 (Part III-a — Part III -- Layers 1-3)
Body location
§III.1 paragraphs 2-3
Substrate
inferential (general ML knowledge)
Claim type
inferential (causal-direction synthesis)
Verdict
Verified

OpenAI red-team paper "OpenAI's approach to external red teaming" — harvest-deferred R1 status

Row
EC-13 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.a closing paragraph
Substrate
explicit-flag (corpus-inventory status; publication-existence + R1 deferred-verbatim)
Claim type
factual (publication-existence) + scope-statement
Verdict
Verified

OpenAI Preparedness Framework + other-frontier-labs analogous documents 2025-2026

Row
EC-19 (Part III-a — Part III -- Layers 1-3)
Body location
§III.3.d paragraph 2
Substrate
inferential (general industry awareness)
Claim type
factual (publication-existence at industry level)
Verdict
Verified

"Multiple lawsuits across 2023-2026 have litigated [training-data IP / fair-use] question"

Row
EC-2 (Part III-a — Part III -- Layers 1-3)
Body location
§III.1 IP paragraph
Substrate
inferential (acknowledged out-of-scope)
Claim type
factual (procedural existence)
Verdict
Verified

Illinois BIPA + state/federal privacy laws apply to training-data PII

Row
EC-3 (Part III-a — Part III -- Layers 1-3)
Body location
§III.1 privacy paragraph
Substrate
inferential (general legal knowledge)
Claim type
factual (statutory-existence)
Verdict
Verified

"*Andreas Horn's ten-layer ladder... omits pre-training data curation entirely*"

Row
EC-4 (Part III-a — Part III -- Layers 1-3)
Body location
§III.1 GRAD-INTERN paragraph 1
Substrate
URL-only — Horn ten-layer LinkedIn ladder; openly credited as the spine origin at Part VIII §VIII.2 and depicted in the §II.3 ecosystem-map figure; comparative omission-claim verifiable against the post
Claim type
factual (compositional claim about Horn's ten-layer enumeration)
Verdict
Verified

RLHF / RLAIF / Constitutional-AI as principal alignment methodologies in 2022-2026 commercial language models

Row
EC-6 (Part III-a — Part III -- Layers 1-3)
Body location
§III.2 technical-issues bullet list
Substrate
inferential (general ML literature)
Claim type
factual (industry-practice-as-described)
Verdict
Verified

Component-part-manufacturer doctrine traditionally applied in non-AI product-liability (automotive component-parts paradigm)

Row
EC-12 (Part III-b — Part III -- Layers 4-7)
Body location
§III.5 ¶3 bullet 1
Substrate
inferential (general tort knowledge)
Claim type
inferential (cross-domain doctrinal-tradition)
Verdict
Verified

The AI-assisted-research hazard: fluent, well-formatted output can be ungrounded; fluency does not indicate grounding

Row
EC-1 (Part IX — Part IX -- Methodology)
Body location
§IX.1
Substrate
inferential (project-authored framing claim)
Claim type
framing (definitional)
Verdict
Verified

The gold / silver / bronze evidence standard; definitions of "warehouse-pinned" and "load-bearing"

Row
EC-2 (Part IX — Part IX -- Methodology)
Body location
§IX.2
Substrate
inferential (project-internal operational taxonomy)
Claim type
definitional (analytical-framework)
Verdict
Verified

The content-addressed store and the five-stage research process

Row
EC-3 (Part IX — Part IX -- Methodology)
Body location
§IX.3
Substrate
inferential (project-internal methodology description)
Claim type
factual (methodological-self-description)
Verdict
Verified

Ground-or-flag and the pre-citation gate, and what each one checks

Row
EC-4 (Part IX — Part IX -- Methodology)
Body location
§IX.4
Substrate
inferential (project-internal methodology description)
Claim type
factual (methodological-self-description)
Verdict
Verified

Worked example — the Section 230 / Lawfare wrong-article catch and its remediation

Row
EC-5 (Part IX — Part IX -- Methodology)
Body location
§IX.4
Substrate
explicit-flag (grounded against the project's own remediation record as conveyed in the Part IX drafting brief; no detail beyond that record is asserted)
Claim type
factual (process event)
Verdict
Verified

The citations index lists every row with tier, status, and chain of custody; the grounding-status figures

Row
EC-6 (Part IX — Part IX -- Methodology)
Body location
§IX.5
Substrate
explicit-flag (grounded against the reconciled citations-page figures of record, 2026-05-22)
Claim type
factual (meta)
Verdict
Verified

The T1–T7 source-type taxonomy

Row
EC-7 (Part IX — Part IX -- Methodology)
Body location
§IX.5
Substrate
inferential (project-internal operational taxonomy)
Claim type
definitional (analytical-framework)
Verdict
Verified

The honest-limits register — six openly-disclosed limitations

Row
EC-8 (Part IX — Part IX -- Methodology)
Body location
§IX.6
Substrate
explicit-flag (project-internal honest-limits register; absorbed from Part VIII §VIII.4.3)
Claim type
scope-statement (research-gap acknowledgment)
Verdict
Verified

§V.5 table cells with *(thin)* / *(follow-on)* markers (8 cells) — explicit flag class

Row
EC-36 (Part V — Part V -- Lab and Platform Responses)
Body location
§V.5 table
Substrate
Claim type
scope-statement
Verdict
Verified

§VI.4 most-understudied "Pre-training data curation literature... is thin in our R1 corpus and is a candidate"

Row
EC-23 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.4 ¶3
Substrate
Claim type
scope-statement
Verdict
Verified

§VI.5 "Watch..." bullets cluster — 8 recommendations-for-attention

Row
EC-24 (Part VI — Part VI -- Strategic Surfaces)
Body location
§VI.5 ¶2-9
Substrate
Claim type
scope-statement
Verdict
Verified

Discovery as channel for accessing operational-texture material in active litigation — doctrinal frontier identified at Part VI §VI.5 "discovery-scope contests" recommendation

Row
EC-12 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.2 attorney-can-do paragraph + closing
Substrate
inferential (cross-reference to Part VI §VI.5 analytical recommendation)
Claim type
inferential (legal-process-doctrine framing)
Verdict
Verified

augmentation harvest retrieval date 2026-05-09; Right to Warn letter integrated into corpus on that date; warehouse pin chunk_hash 437dd326c0…

Row
EC-13 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.0 + §VII.1
Substrate
explicit-flag
Source
https://righttowarn.ai/
Claim type
factual (meta)
Verdict
Provisional

Lab-departure interview channel (former-OpenAI safety-team-member departure interview) attempted at but not retrieved at access date; channel remains unaddressed in our corpus

Row
EC-14 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.0 + §VII.1 "what we do not have" + §VII.2 augmentation summary
Substrate
explicit-flag (project-internal augmentation negative-result; documented absence)
Claim type
scope-statement (research-gap acknowledgment)
Verdict
Verified

§VII.2 source-type-by-source-type augmentation summary (7 categories: whistleblower-disclosure substantive; long-form lab-leadership supplementary; interview-format lab-departure attempted-not-retrieved; specialist podcasts front-page-only; academic/think-tank publication-index-only; trade-press niche front-page-only; LinkedIn-tier dept-org partial)

Row
EC-15 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.2 augmentation summary bullet-list
Substrate
explicit-flag (project-internal augmentation-manifest meta-claim)
Claim type
scope-statement (research-corpus characterization)
Verdict
Provisional

§VIII.4.3 honest-limits register Q14 entry "partial-with-substantive-whistleblower-channel-evidence" cross-reference

Row
EC-16 (Part VII — Part VII -- Coordination Gaps)
Body location
§VII.0 + §VII.1 + closing
Substrate
inferential (project-internal cross-reference to Part VIII §VIII.4.3 honest-limits register)
Claim type
scope-statement (cross-reference consistency)
Verdict
Verified

"4,101 SHA-hashed sentence-level pins drawn from 43 sources"

Row
EC-1 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.1
Substrate
explicit-flag
Claim type
factual (meta)
Verdict
Provisional

§VIII.3.1 enumeration of 43 R1 primary-source-pinned sources

Row
EC-11 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.1
Substrate
explicit-flag (project-internal corpus-manifest meta-claim -- 43 R1 sources)
Claim type
factual (meta)
Verdict
Provisional

§VIII.3.2 19+ cross-corpus pinned-eligible sources (Conway + Brandenburg + Counterman + 16+ academic + statutes)

Row
EC-12 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.2
Substrate
inferential (project-internal cross-corpus source enumeration)
Claim type
factual (citation list)
Verdict
Verified

§VIII.3.2 Goodwin removed 2026-05-13 per Q-MATRIX-2 ratification

Row
EC-13 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.3.2 + Part IV §IV.1 + §IV.3.d Goodwin cuts
Substrate
explicit-flag (project-internal decision record -- Goodwin cut per Q-MATRIX-2)
Claim type
factual (in-corpus mutation)
Verdict
Verified

§VIII.5 glossary terms (16 definitions)

Row
EC-18 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.5
Substrate
inferential (project-authored glossary definitions)
Claim type
definitional (per term)
Verdict
Verified

"Pinned source" definition at §VIII.5 ("source processed by the research pipeline with a SHA256 hash on its sentence-level content")

Row
EC-19 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.5
Substrate
inferential (project-internal operational definition)
Claim type
definitional
Verdict
Verified

§VIII.2 per-layer fact sheets (7 layers L1-L7) — each with primary citations

Row
EC-8 (Part VIII — Part VIII -- Appendices)
Body location
§VIII.2.L1-L7
Substrate
WAREHOUSE-PINNED [multi-source: per-layer L1-L7 fact-sheet substrate -- RSP v3 + OpenAI Safety Hub + CISA TEVV + McGuireWoods + Psychiatric Times]
Claim type
factual (per-layer reference)
Verdict
Verified