{ "format": "agent-article-v0", "article_id": "EA-040", "canonical_url": "/tools/ea-040-extension-research-run-2026-09-28-16-19-jst", "page_type": "reference", "claim_ids": [], "evidence_ids": [ "EVD-EA-040-URL-1106781EBBDE", "EVD-EA-040-URL-E639F95D9C53", "EVD-EA-040-URL-D41993CA6F1C", "EVD-EA-040-URL-33D25B306E1A", "EVD-EA-040-URL-71C6BF63EBCB", "EVD-EA-040-URL-49AF6AFC8C20", "EVD-EA-040-URL-D7AA87223273", "EA-040-R11-RESEARCH", "EA-040-R11-WEB-READBACK" ], "unknown_ids": [ "EA-040-UNKNOWN-01", "EA-040-UNKNOWN-CURRENTNESS" ], "operational_object_ids": [], "provider": [], "models": [], "plans": [], "surfaces": [], "regions": [], "current_status": "UNKNOWN", "observation_status": "SPEC_ONLY", "freshness": "UNKNOWN", "last_verified": null, "supersedes": [], "superseded_by": [], "provenance": [ "LIB-EA0040-20260928", "SRC-EA-040-URL-1106781EBBDE", "SRC-EA-040-URL-E639F95D9C53", "SRC-EA-040-URL-D41993CA6F1C", "SRC-EA-040-URL-33D25B306E1A", "SRC-EA-040-URL-71C6BF63EBCB", "SRC-EA-040-URL-49AF6AFC8C20", "SRC-EA-040-URL-D7AA87223273", "WEBBODY-EA0040-R8-7EB75A8A" ], "article_sources": [ { "id": "EVD-EA-040-URL-1106781EBBDE", "evidence_class": "RESEARCH", "title": "2026-09-28,", "href": "https://www.iso.org/standard/74296.html", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-1106781EBBDE" ], "supporting_locators": [ "EVIDENCE TABLE", "2026-09-28, https://www.iso.org/standard/74296.html" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EVD-EA-040-URL-E639F95D9C53", "evidence_class": "RESEARCH", "title": "access 2026-09-28,", "href": "https://www.iso.org/standard/74438.html", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-E639F95D9C53" ], "supporting_locators": [ "EVIDENCE TABLE", "access 2026-09-28, https://www.iso.org/standard/74438.html" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EVD-EA-040-URL-D41993CA6F1C", "evidence_class": "RESEARCH", "title": "https://www.iso.org/standard/81230.html", "href": "https://www.iso.org/standard/81230.html", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-D41993CA6F1C" ], "supporting_locators": [ "EVIDENCE TABLE", "https://www.iso.org/standard/81230.html" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EVD-EA-040-URL-33D25B306E1A", "evidence_class": "RESEARCH", "title": "2026-09-28,", "href": "https://doi.org/10.6028/NIST.AI.600-1", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-33D25B306E1A" ], "supporting_locators": [ "EVIDENCE TABLE", "2026-09-28, https://doi.org/10.6028/NIST.AI.600-1" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EVD-EA-040-URL-71C6BF63EBCB", "evidence_class": "RESEARCH", "title": "access 2026-09-28,", "href": "https://airc.nist.gov/", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-71C6BF63EBCB" ], "supporting_locators": [ "EVIDENCE TABLE", "access 2026-09-28, https://airc.nist.gov/" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EVD-EA-040-URL-49AF6AFC8C20", "evidence_class": "RESEARCH", "title": "https://opentelemetry.io/docs/specs/otel/semantic-conventions/", "href": "https://opentelemetry.io/docs/specs/otel/semantic-conventions/", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-49AF6AFC8C20" ], "supporting_locators": [ "EVIDENCE TABLE", "https://opentelemetry.io/docs/specs/otel/semantic-conventions/" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EVD-EA-040-URL-D7AA87223273", "evidence_class": "RESEARCH", "title": "https://genai.owasp.org/2026/09/01/owasp-genai-security-project-unveils-2026-top-10-for-llm-applications-new-agent-control-standard-and-sponsors-as-community-tops-30000-members/", "href": "https://genai.owasp.org/2026/09/01/owasp-genai-security-project-unveils-2026-top-10-for-llm-applications-new-agent-control-standard-and-sponsors-as-community-tops-30000-members/", "supports": [], "limit": "元の監査済み資料に記載された参照先です。URLの復元は、資料の最新性や全主張の支持を新たに検証したことを意味しません。", "origin_class": "SOURCE_PACKET_REFERENCE", "source_or_experiment_ids": [ "SRC-EA-040-URL-D7AA87223273" ], "supporting_locators": [ "EVIDENCE TABLE", "https://genai.owasp.org/2026/09/01/owasp-genai-security-project-unveils-2026-top-10-for-llm-applications-new-agent-control-standard-and-sponsors-as-community-tops-30000-members/" ], "scope": "EVIDENCE TABLE", "provenance_status": "SOURCE_PACKET_REFERENCE_PROJECTED" }, { "id": "EA-040-R11-RESEARCH", "evidence_class": "RESEARCH", "title": "EA-040 監査済みresearch / canonical revision binding", "supports": [ "F1 \\[RESEARCH\\]: 単一の引用元がEnterprise AI\nCoreをend-to-endで規定しているわけではない。防御可能な構成は複数authorityのcrosswalk/profileであり、「普遍的なEnterprise\nAI Core標準が既に存在する」とは記述しない。" ], "limit": "canonical監査・revision bindingの存在を示す。外部ソースcurrentnessが本文またはcontrol-planeで確定していない場合は、currentnessをUNKNOWNのまま保持する。", "origin_class": "AUDITED_CANONICAL_RESEARCH_PACKET", "source_or_experiment_ids": [ "LIB-EA0040-20260928" ], "supporting_locators": [ "LIB-EA0040-20260928", "EA-040-R5-LEGACY-MIGRATION-20261008-R1", "content/staging/EA-040.md" ], "scope": "監査済みEA本文・canonical revision・既存Source Graph relationの範囲。欠損currentnessは推測で閉じない。", "provenance_status": "AUDIT_PASS_CANONICAL_BINDING" }, { "id": "EA-040-R11-WEB-READBACK", "evidence_class": "IMPL", "title": "EA-040 Web body identity / canonical readback", "supports": [ "canonical staging bodyのSHA-256一致とR9 canonical ledger readbackをProduction projectionの入力として検証する。" ], "limit": "body identity/readbackは本文claim自体の真偽や外部ソースの最新性を追加保証しない。", "checked_at": null, "origin_class": "IMPLEMENTATION_READBACK", "source_or_experiment_ids": [ "WEBBODY-EA0040-R8-7EB75A8A" ], "supporting_locators": [ "WEBBODY-EA0040-R8-7EB75A8A", "content/staging/EA-040.md", "EA-040-R5-LEGACY-MIGRATION-20261008-R1" ], "scope": "R9/R10 canonical readback後のProduction生成時body identity。", "provenance_status": "CANONICAL_READBACK_BODY_HASH_MATCH" } ], "research_sections": [ { "heading": "要点と前提", "text": "RUN_MODE: LEDGER_DRIVEN_RESEARCH + CANONICAL_SYNC_RESCUE LEDGER_ID:\n1KqhJtiSb2S84rWmeYVwKcT_76dBwZ_B0ZZuAWQqH0rQ SYNC_RESCUE: EA-211, EA-212\nlocation-audit ledger writeback completed and read back before research\nunit." }, { "heading": "WHY_THIS_MATTERS_TO_UPDATE", "text": "製品名・provider固有UI・model名を正本概念にすると更新のたびに運用仕様が崩れる。Coreを能力、境界、状態、証拠、統制点として定義すれば、provider\nmappingを差替可能にしつつ監査・移行・比較の共通面を維持できる。" }, { "heading": "SCOPE_ROUTING", "text": "COMMON_CORE / PROVIDER_NEUTRAL /\nGOVERNANCE+OPERABILITY+OBSERVABILITY。provider固有実装、特定tenant挙動、契約・plan\nentitlementはCore定義から分離する。" }, { "heading": "CLAIM_OR_RESEARCH_QUESTION", "text": "Enterprise AI\nCoreを単一の新標準として発明せず、既存のAI用語・システム枠組み・管理システム・リスク管理・TEVV・observability/security\nconventionsをcrosswalkし、最小のvendor-neutral contractを定義できるか。" }, { "heading": "EVIDENCE TABLE", "text": "1. official standard --- ISO/IEC 22989:2022, *Information technology\n --- Artificial intelligence --- Artificial intelligence concepts and\n terminology*, ISO/IEC JTC 1/SC 42, 2022-07, Edition 1, access\n 2026-09-28, https://www.iso.org/standard/74296.html, VERIFIED_SPEC,\n terminology baseline. AI\n terminology/conceptsを組織横断で提供するがenterprise runtime\n Coreは規定しない。\n2. official standard --- ISO/IEC 23053:2022, *Framework for AI systems\n using machine learning*, ISO/IEC JTC 1/SC 42, 2022-06, Edition 1,\n access 2026-09-28, https://www.iso.org/standard/74438.html,\n VERIFIED_SPEC, system/component baseline. ML-based AI\n systemを構造的に記述する枠組みでありagent/RAG/provider-neutral\n operational contractそのものではない。\n3. official standard --- ISO/IEC 42001:2023, *Information technology\n --- Artificial intelligence --- Management system*, ISO/IEC JTC 1/SC\n 42, 2023-12, Edition 1, access 2026-09-28,\n https://www.iso.org/standard/81230.html, VERIFIED_SPEC,\n organizational governance boundary. AI management\n systemの確立・実装・維持・継続改善を対象とする。\n4. government framework --- NIST AI 600-1, *Artificial Intelligence\n Risk Management Framework: Generative Artificial Intelligence\n Profile*, NIST, 2024-07-26, NIST page updated 2026-04-08, access\n 2026-09-28, https://doi.org/10.6028/NIST.AI.600-1, VERIFIED_SPEC,\n risk/lifecycle/evidence controls. AI RMF 1.0 companion\n profileであり製品保証ではない。\n5. government operational resource --- NIST AI Resource Center, NIST,\n access 2026-09-28, https://airc.nist.gov/, VERIFIED_SPEC, TEVV\n operationalization. Testing/evaluation/verification/validation\n resourcesを提供する。\n6. author-maintained OSS specification --- OpenTelemetry Semantic\n Conventions, OpenTelemetry, access 2026-09-28,\n https://opentelemetry.io/docs/specs/otel/semantic-conventions/,\n status=Development on retrieved spec, OSS_PRACTICE, portable\n telemetry vocabulary. Convention\n maturityは個別に異なるためversion/maturityを保持する。\n7. security community primary project --- OWASP GenAI Security Project\n 2026 Top 10 / Agent Control Standard announcement, OWASP GenAI\n Security Project, 2026-09-01, access 2026-09-28,\n https://genai.owasp.org/2026/09/01/owasp-genai-security-project-unveils-2026-top-10-for-llm-applications-new-agent-control-standard-and-sponsors-as-community-tops-30000-members/,\n SOURCE_REPORT, security/failure-envelope\n input。ISO/NIST相当の普遍的normative standardとは扱わない。" }, { "heading": "FINDINGS", "text": "F1 \\[RESEARCH\\]: 単一の引用元がEnterprise AI\nCoreをend-to-endで規定しているわけではない。防御可能な構成は複数authorityのcrosswalk/profileであり、「普遍的なEnterprise\nAI Core標準が既に存在する」とは記述しない。\n\nF2 \\[PROPOSAL\\]: 最小Core objectは六planeに分離する。(A) CAPABILITY ---\ninference/retrieval/memory/tool execution/media/eval、(B) BOUNDARY ---\nidentity/authorization/data/tenant/network/tool trust boundary、(C)\nSTATE ---\nmodel/config/prompt/policy/index/memory/session/workflow/version\nstate、(D) EVIDENCE --- provenance/source\nlocator/eval/trace/log/audit/readback、(E) CONTROL --- policy\ngate/approval/budget/rate/cost/rollback/revocation、(F) LIFECYCLE ---\ndesign/build/deploy/operate/change/retire。これは本辞典のsynthesisであり外部標準の引用条項ではない。\n\nF3 \\[PROPOSAL\\]: provider mappingはprovider objectをCore capability\nID、availability gate、evidence locatorへbindする。mappingからCore\nsemanticsを逆定義しない。未確認tenant挙動はUNKNOWNのまま保持する。\n\nF4 \\[RESEARCH\\]: ISO/IEC 42001は組織management-system、NIST AI RMF/GAI\nProfileはrisk-management outcomes/actions、NIST AIRCはTEVV\noperationalization、OpenTelemetryはportable telemetry\nnamingを主に担う。相互補完であり代替関係ではない。\n\nF5 \\[PRODUCT_VOLATILE\\]: NISTはAI RMF 1.0を2026年に改訂中と明記。ISO/IEC\n22989 GenAI Amendment 1とISO/IEC 23053 GenAI Amendment\n1は2026-09-28時点でunder development/final-draft\napproval。draftをpublished standardとして扱わない。" }, { "heading": "LIMITS / COUNTEREVIDENCE / FAILURE ENVELOPE", "text": "- terminology、management system、risk\n framework、observability、security guidanceはconformance\n semanticsが異なる。名称類似だけでcrosswalk equivalenceを断定しない。\n- ISO/IEC 42001への適合・認証は特定agent/tool call、retrieval\n ACL、memory deletion、model responseの正しさを証明しない。\n- OpenTelemetry convention準拠は相関可能性を改善するが、telemetry\n completeness、truthfulness、tenant isolation、audit-grade\n retentionを自動的に証明しない。\n- model ID、API、plan、region、policy\n controlは標準より速く変化する。provider\n mappingにはas-of/version/readbackが必要。\n- public docsやmarketing pageをOBSERVED tenant\n behaviorへ昇格させない。\n- agentic systemではmodel\n outputがactionへ変換されるため、identity/authorization/tool/data\n boundaryへcontrolを結び付ける。\n- lowest-common-denominator schemaはprovider固有control/failure\n modeを消し得るためextension fieldとprovider-specific\n sectionを許容する。" }, { "heading": "AVAILABILITY_GATES", "text": "Core specification自体はprovider-neutral。product\nmappingではprovider、product/API、model/version、plan/contract、region、organization/tenant\npolicy、identity/role、data residency、feature flag/preview\nstatus、dateを分離記録する。organizationのcertification/framework\nadoptionは別gateであり推測しない。" }, { "heading": "IMPLEMENTATION_STATE", "text": "PROPOSAL_READY_FOR_COMPILATION。外部の単一標準ではない。構成層は既存authorityで支持されるが、辞典独自Core\nschema/profileにはschema design、conformance rule、provider mapping\ntestが必要。認証済みtenant操作は未実施。" }, { "heading": "BENCHMARK_OR_EVAL", "text": "benchmark scoreだけでCore\nconformanceを証明しない。将来TEST_PROFILEは、少なくとも2 providerと1\nOSS/local stackで同一bounded workflowをCore\nobjectへmapし、identity/authorization boundary、retrieval\nprovenance、state/version capture、tool action evidence、cost/usage\ntelemetry、failure/rollback、readbackを検証する。mapping\ncompleteness、evidence reproducibility、model task\nqualityは別scoreにする。NIST AIRC/TEVVはmethodology\ninputであり本Coreのpass/fail benchmarkではない。" }, { "heading": "SECURITY", "text": "capability invocationはactor/workload identity、delegated\nauthority、target resource/tool、policy decisionをbindする。model\noutput単独をauthorizationにしない。 PRIVACY: data\nclass、purpose/retention/residency constraint、provider\ntraining/retention control\napplicabilityを記録し、tenant実設定は別検証する。 AUDIT: event\ncorrelation、source/provenance locator、config/model/policy\nversion、result/readback relationを保持する。telemetryとaudit\nevidenceは別evidence classとする。 COST: metering unit、price\nsource/as-of、budget/rate gate、observed usageを分離する。public list\npriceはorganization TCOではない。 OPERABILITY: health/dependency\nstate、retry/idempotency、rollback/revocation、degradation\nmodeを記録し、provider非公開内部挙動はUNKNOWNとする。" }, { "heading": "SAFE_WORDING_FOR_FUTURE_BOOK", "text": "「Enterprise AI\nCore」は既存の単一国際標準の名称ではなく、本辞典がベンダー中立の運用比較面として定義するprofileである。ISO/IEC\n22989・23053・42001、NIST AI RMF/GAI\nProfile、TEVV資源、OpenTelemetry等の既存標準・枠組みを置換せず、能力・境界・状態・証拠・統制・ライフサイクルをcrosswalkする。provider固有機能とtenant実挙動は別レイヤでas-of/version/availability\ngate付きで検証する。" }, { "heading": "UPDATE_WORK_HANDOFF", "text": "1. Core schemaを capability/boundary/state/evidence/control/lifecycle\n の六planeで定義する。\n2. 各fieldに normative source / dictionary proposal / provider\n extension のoriginを保持する。\n3. conformanceを Core schema completeness / evidence reproducibility /\n provider mapping fidelity に分離する。\n4. NIST AI RMF revisionおよびISO GenAI\n amendments公開後にVERSION_REFRESHする。\n5. provider\n mappingをCoreへ従属させ、provider差分からCoreを逆定義しない。" }, { "heading": "UNKNOWN", "text": "ISO GenAI amendmentsの最終公開内容・日付、NIST AI RMF\nrevision最終内容、OpenTelemetry GenAI\nconventions各要素のstable/experimental状態、この六plane\nsynthesisそのものの独立enterprise実装evidence、organization-specific\ncertification/control mapping。" }, { "heading": "STOP_CONDITION", "text": "EA-040 bounded\nresearchは、単一Core標準が直接存在するとは確認できないこと、crosswalk可能な主要層、version-sensitive\ndependencies、failure\nenvelope、安全な将来記述を確定したためCOMPLETE。次回は再研究せず、標準改訂時のVERSION_REFRESHまたは実装検証時のSUPPLEMENT_ONLYとする。\n\n---" }, { "heading": "AUDITED_BOUNDARY", "text": "Six-plane Enterprise AI Core synthesis is DICTIONARY PROPOSAL, not an existing universal standard. Preserve standards/tenant boundaries and every OPEN_UNKNOWN. NIST AI RMF revision, ISO GenAI amendments, and OpenTelemetry GenAI maturity remain version-sensitive; no draft is represented as published.\nTESTS: Library authority full readback PASS; proposal/standards-boundary/UNKNOWN markers PASS. Build/CI deferred to eligible Git batch." } ], "human_url": "/tools/ea-040-extension-research-run-2026-09-28-16-19-jst", "markdown_url": "/agent/article.md?path=%2Ftools%2Fea-040-extension-research-run-2026-09-28-16-19-jst", "json_url": "/agent/article.json?path=%2Ftools%2Fea-040-extension-research-run-2026-09-28-16-19-jst" }