米国 NIST IR 8596(初期ドラフト)人工知能向けサイバーセキュリティ枠組みプロファイル(Cyber AI Profile):NISTコミュニティプロファイル (2025.12.16)
こんにちは、丸山満彦です。
NISTが、IR 8596(初期ドラフト)人工知能向けサイバーセキュリティ枠組みプロファイル(Cyber AI Profile):NISTコミュニティプロファイル
を公表していますね...
CSF2.0をベースにさまざまなガイドラインを組み立てられるので、CSF2.0を策定した意味は非常に大きいですね...日本では、ISMS制度のクライテリアの一部にもなっているJISQ27001(実践規範としての JISQ27002)があったはずですが、サイバー攻撃に対してはCSF2.0がわかりやすいのと、日本も含めてグローバルな金融機関が標準的に利用しているCRI ProfileにもCSF2.0がベースとなっているし、経済産業省のサプライチェーン評価制度についてもCSSF2.0がベースになっているので、これから徐々にJISQ27001、JISQ27002からCSF2.0に移っていくのですかね...
ただ、CSF2.0は環境的な脅威、故障といった脅威に対するコントロールが十分ではないようにもみえるのでちょっと注意をしていたほうがよいと思います...
さて、AIとセキュリティですが、、、
基本的には
・AIシステムに対するセキュリティ
・AIシステムを利用した防御
・AIシステムを利用した攻撃の阻止
ということで、私が2020年に白浜のシンポジウムで話をした視点と同じですかね...
● NIST - ITL
| NIST IR 8596 (Initial Preliminary Draft) Cybersecurity Framework Profile for Artificial Intelligence (Cyber AI Profile): NIST Community Profile | NIST IR 8596(初期ドラフト)人工知能向けサイバーセキュリティ枠組みプロファイル(Cyber AI Profile):NISTコミュニティプロファイル |
| Announcement | お知らせ |
| Check out NIST’s Cyber AI Profile Preliminary Draft and Save the Date for our Cyber AI Workshop #2 in January | NISTのCyber AI Profile初期ドラフトを確認し、1月開催のCyber AIワークショップ#2の日程を予定にいれてください |
| Draft for Public Comment | 公開コメント用ドラフト |
| AI has become a driving force behind today’s technological development, transforming industries and redefining how society operates. Advancements in AI technology introduce both cybersecurity opportunities and challenges to organizations. | AIは今日の技術開発の原動力となり、産業を変革し、社会の運営方法を再定義している。AI技術の進歩は、組織にとってサイバーセキュリティの機会と課題の両方をもたらす。 |
| NIST’s preliminary draft Cyber AI Profile can help organizations strategically adopt AI while addressing and prioritizing cybersecurity risks stemming from its advancements. | NISTの初期ドラフト「サイバーAIプロファイル」は、組織がAIを戦略的に導入すると同時に、その進歩に伴うサイバーセキュリティリスクに対処し優先順位をつけるのに役立つ。 |
| The Cyber AI Profile addresses the following Focus Areas: | サイバーAIプロファイルは以下の重点領域に対処する: |
| 1. Securing AI System Components (Secure) | 1. AIシステムコンポーネントの保護(Secure) |
| 2. Conducting AI-Enabled Cyber Defense (Defend) | 2. AIを活用したサイバー防御の実施(Defend) |
| 3. Thwarting AI-enabled Cyber Attacks (Thwart) | 3. AIを活用したサイバー攻撃の阻止(Thwart) |
| Abstract | 概要 |
| The Cybersecurity Framework Profile for Artificial Intelligence (AI) Profile (“Cyber AI Profile” or “The Profile”) will provide guidelines for managing cybersecurity risk related to AI systems as well as identifying opportunities for using AI to enhance cybersecurity capabilities. The Profile is organized using the NIST Cybersecurity Framework 2.0 outcomes (Functions, Categories, and Subcategories). This Preliminary Draft is shared along with a request for public comment to solicit feedback on the planned direction and content. Comments received will inform the initial public draft. More information about this project, including a roadmap, is available on the National Cybersecurity Center of Excellence (NCCoE) Cyber AI Profile project page. | 人工知能(AI)向けサイバーセキュリティ枠組みプロファイル(「サイバーAIプロファイル」または「本プロファイル」)は、AIシステムに関連するサイバーセキュリティリスクの管理指針を提供するとともに、AIを活用してサイバーセキュリティ能力を強化する機会を識別するものである。本プロファイルは、NISTサイバーセキュリティフレームワーク2.0の成果物(機能、カテゴリー、サブカテゴリー)を用いて構成されている。この初期ドラフトは、計画された方向性と内容に関するフィードバックを求めるため、パブリックコメントの募集と共に共有される。寄せられたコメントは、最初の公開草案の作成に反映される。本プロジェクトに関する詳細情報(ロードマップを含む)は、国立サイバーセキュリティ・センター・オブ・エクセレンス(NCCoE)のサイバーAIプロファイルプロジェクトページで入手可能である。 |
| Notes for Reviewers | 査読者向け注意事項 |
| The purpose of this initial draft is to share insights regarding the technical content direction of the Cyber AI Profile. NIST welcomes feedback and opinions on all aspects of this publication. Specifically, NIST seeks views on the following points: | 本ドラフトの目的は、サイバーAIプロファイルの技術的内容の方向性に関する知見を共有することである。NISTは本出版物のあらゆる側面に関するフィードバックや意見を歓迎する。特にNISTは以下の点について見解を求めている: |
| 1. Document structure and subject matter: | 1. 文書構造と主題内容: |
| a) How do you intend to utilise this publication? What changes would you like to see to increase or improve its utilisation? | a) 本出版物をどのように活用する予定か?その活用度を高めたり改善したりするために、どのような変更を希望するか? |
| b) How do you anticipate this publication will influence future practices or processes? | b) 本出版物が将来の実践やプロセスにどのような影響を与えると予想するか? |
| c) Are the proposed topics in this document sufficient for organisations to prioritise their AI cybersecurity outcomes? | c) 本文書で提案されているトピックは、組織がAIサイバーセキュリティ成果を優先順位付けするのに十分か? |
| 2. Description of Focus Areas (Section 2.1): | 2. 重点領域の説明(セクション2.1): |
| a) To what extent does the description of focus areas reflect the scope and characteristics of AI usage? Are any characteristics missing? If so, what are they? How should they be described? | a) 重点領域の説明は、AI利用の範囲と特性をどの程度反映しているか? 欠落している特性はあるか? ある場合、それは何か? どのように記述すべきか? |
| 3. Profile Content (Sections 2.3–2.8): | 3. プロファイル内容(セクション2.3~2.8): |
| a) When considering the application of the Cyber AI Profile, how useful (or useless) is it to display the three focus areas in parallel (the current presentation method)? What value is there in providing the profile content separately for each focus area? | a) サイバーAIプロファイルの適用を検討する際、3つの重点領域を並列表示する(現在の提示方法)ことはどの程度有用(または無用)か?各重点領域ごとにプロファイル内容を個別に提供することにはどのような価値があるか? |
| b) What format would be useful for providing the Cyber AI Profile information (e.g., spreadsheet/workbook, NIST Cybersecurity and Privacy Reference Tool (CPRT))? | b) サイバーAIプロファイル情報を提供するのに有用な形式は何か(例:スプレッドシート/ワークブック、NISTサイバーセキュリティ・プライバシー参照ツール(CPRT))? |
| c) To what extent are the priorities and considerations discussed in Sections 2.3-2.8 relevant to existing practices or standards utilised by your organisation? | c) セクション2.3~2.8で議論された優先事項と考慮事項は、組織が利用する既存の実践や標準にどの程度関連しているか? |
| Are there significant gaps between current practices and the practices needed to address the unique characteristics of AI within each focus area that this publication should address? How should AI-specific considerations be reflected in the prioritisation of each subcategory? | 現在の実践と、本出版物が対処すべき各重点領域におけるAIの固有特性を扱うために必要な実践との間に、重大な隔たりはあるか? AI固有の考慮事項は、各サブカテゴリーの優先順位付けにどう反映されるべきか? |
| d) NIST has published reference materials | d) NISTはサイバーセキュリティ枠組み(CSF)2.0の参考資料 |
| and implementation examples for the Cybersecurity Framework (CSF) 2.0, illustrating potential approaches to achieving outcomes within each subcategory. This initial draft includes examples of informational references for the Cyber AI Profile. Literature review is ongoing, and NIST seeks further input on additional informational references to include. What additional AI cybersecurity guidelines, standards, best practices, or mappings would you recommend adding as supporting references for the Cyber AI Profile? For any recommended supporting references, please share the rationale and how and why they should be prioritised within this document. | 及び実装例を公開しており、各サブカテゴリーにおける成果達成の潜在的なアプローチを示している。本草案にはサイバーAIプロファイルの情報的参照例が含まれる。文献調査は継続中であり、NISTは追加情報的参照の提案を求めている。サイバーAIプロファイルの補足参照として追加すべきAIサイバーセキュリティガイドライン、標準、ベストプラクティス、マッピングを推奨するものはあるか?推奨する補足参照資料については、その根拠と、本文書内で優先すべき理由及び方法を共有してほしい。 |
| 4. Glossary (Annex B): | 4. 用語集(附属書B): |
| a) NIST welcomes requests and suggestions for terms to add to the glossary in this document. | a) NISTは、本文書の用語集に追加すべき用語の要望や提案を歓迎する。 |
| Comment submitters are encouraged to use the comment template provided on the NCCoE’s Cyber AI Profile project page for responses to the question set and specific comments on the document text. Feedback and completed comment templates may also be submitted to the project team at cyberaiprofile@nist.gov. The comment submission deadline is 30 January 2026 at 11:59 PM Eastern Time. | コメント提出者は、質問セットへの回答及び文書本文への具体的なコメントについて、NCCoEのサイバーAIプロファイルプロジェクトページで提供されているコメントテンプレートの使用を推奨される。フィードバック及び完成したコメントテンプレートは、プロジェクトチーム宛てに cyberaiprofile@nist.gov へ提出することも可能である。コメント提出期限は、東部時間 2026年1月30日 23:59 である。 |
| All comments may be made public under the Freedom of Information Act. | 全てのコメントは、情報公開法に基づき公開される可能性がある。 |
・[PDF] IR.8596.iprd
目次...
| Executive Summary | エグゼクティブサマリー |
| 1. Introduction | 1. 序論 |
| 1.1. Purpose | 1.1. 目的 |
| 1.2. Scope | 1.2. 範囲 |
| 1.3. Target Audience | 1.3. 対象読者 |
| 1.4. Document Structure | 1.4. 文書構成 |
| 2. Cyber AI Profile | 2. サイバーAIプロファイル |
| 2.1. Focus Areas | 2.1. 重点領域 |
| 2.1.1. Secure AI System Components | 2.1.1. セキュアなAIシステム構成要素 |
| 2.1.2. Implementing AI-Enabled Cyber Defence (Defend) | 2.1.2. AIを活用したサイバー防衛の実装(Defend) |
| 2.1.3. Thwarting AI-Enabled Cyber Attacks (Thwart) | 2.1.3. AIを活用したサイバー攻撃の阻止(Thwart) |
| 2.2. How to Read the Cyber AI Profile | 2.2. サイバーAIプロファイルの読み方 |
| 2.3. Cyber AI Profile: Governance | 2.3. サイバーAIプロファイル:統治 |
| 2.4. Cyber AI Profile: Identify | 2.4. サイバーAIプロファイル:識別 |
| 2.5. Cyber AI Profile: Protect | 2.5. サイバーAIプロファイル:防御 |
| 2.6. Cyber AI Profile: Detect | 2.6. サイバーAIプロファイル:検知 |
| 2.7. Cyber AI Profile: Respond | 2.7. サイバーAIプロファイル:対応 |
| 2.8. Cyber AI Profile: Recover | 2.8. サイバーAIプロファイル:復旧 |
| References | 参考文献 |
| Annex A. List of Symbols, Abbreviations and Acronyms | 附属書A. 記号、略語、頭字語一覧 |
| Annex B. Glossary | 附属書B. 用語集 |
| Annex C. Cybersecurity Framework 2.0 Overview | 附属書C. サイバーセキュリティフレームワーク2.0概要 |
| Annex D. How to Use the Cyber AI Profile | 附属書D. サイバー AI プロファイルの使用方法 |
| List of Tables | 表一覧 |
| Table 1 Cyber AI Profile – GOVERN | 表1 サイバー AI プロファイル – 統治 |
| Table 2 Cyber AI Profile – Identify | 表2 サイバー AI プロファイル – 識別 |
| Table 3 Cyber AI Profile – Defend | 表3 サイバー AI プロファイル – 防御 |
| Table 4 Cyber AI Profile – Detect | 表4 サイバー AI プロファイル – 検知 |
| Table 5 Cyber AI Profile – Respond | 表5 サイバー AI プロファイル – 対応 |
| Table 6 Cyber AI Profile – RECOVER | 表6 サイバー AI プロファイル – 回復 |
| List of Figures | 図一覧 |
| Figure 1: Relationships Between Cyber AI Profile Focus Areas | 図1: サイバー AI プロファイルの重点領域間の関係 |
| Figure 2: How to Read the Cyber AI Profile Initial Draft Tables | 図2:サイバーAIプロファイル初期草案表の読み方 |
| Figure 3: CSF Core Structure | 図3:CSFコア構造 |
| Executive Summary | エグゼクティブサマリー |
| Recent advances in artificial intelligence (AI) technology present new cybersecurity opportunities and challenges for organisations. NIST, with its longstanding activities in both cybersecurity and AI, has worked closely with stakeholders in conducting research and advancing standards and technologies. | 人工知能(AI)技術の最近の進歩は、組織にとって新たなサイバーセキュリティの機会と課題をもたらしている。NISTは、サイバーセキュリティとAIの両分野における長年の活動を通じて、研究の実施や標準・技術の推進において関係者と緊密に連携してきた。 |
| NIST is leveraging its cybersecurity and AI resources to help organisations address these new AI opportunities and challenges. Working with the cybersecurity and AI communities, NIST is developing the ‘Cybersecurity Framework Profile for Artificial Intelligence’ (the ‘Cyber AI Profile’ or ‘The Profile’). | NISTは、組織がこれらの新たなAIの機会と課題に対処するのを支援するため、サイバーセキュリティとAIのリソースを活用している。サイバーセキュリティとAIコミュニティと連携し、NISTは「人工知能向けサイバーセキュリティ枠組みプロファイル」(「サイバーAIプロファイル」または「本プロファイル」)を開発中である。 |
| This aims to establish a consistent, common approach for discussing the impact of AI advancements on an organisation's cybersecurity risk management strategy, whilst providing resources for organisations to strategically adopt AI, addressing and prioritising new cybersecurity risks that accompany AI progress. The Cyber AI Profile does not replace existing frameworks but provides prioritised cybersecurity guidance for organisations working to protect AI, enhance AI-enabled cybersecurity defences, and defend against AI misuse. It leverages the capabilities, categories, and subcategories of CSF 2.0 to present informed, structured, and technology-neutral recommendations for cybersecurity and AI professionals. | これは、AIの進歩が組織のサイバーセキュリティリスクマネジメント戦略に与える影響を議論するための一貫した共通アプローチを確立すると同時に、組織が戦略的にAIを導入し、AIの進展に伴う新たなサイバーセキュリティリスクに対処・優先順位付けするためのリソースを提供することを目的とする。サイバーAIプロファイルは既存の枠組みに取って代わるものではなく、AIの防御、AIを活用したサイバーセキュリティ防御の強化、AIの悪用に対する防御に取り組む組織向けに優先順位付けされたサイバーセキュリティガイダンスを提供する。CSF 2.0の能力、カテゴリー、サブカテゴリーを活用し、サイバーセキュリティおよびAI専門家向けに情報に基づいた構造化された技術中立的な推奨事項を提示する。 |
| The Cyber AI Profile addresses three key focus areas: | サイバーAIプロファイルは3つの主要な重点領域に対処する: |
| • Securing AI System Components: Focuses on identifying cybersecurity challenges when integrating AI into an organisation's ecosystem or infrastructure. | • AIシステムコンポーネントの保護:組織のエコシステムやインフラにAIを統合する際のサイバーセキュリティ課題の特定に焦点を当てる。 |
| • Implementing AI-Enabled Cyber Defences (Defend): Focuses on identifying opportunities to leverage AI to enhance cybersecurity processes and activities, and understanding the challenges of using AI to support defensive operations. | • AIを活用したサイバー防御の実装(防御):サイバーセキュリティプロセスや活動を強化するためにAIを活用する機会の特定、および防御活動を支援するAI利用の課題理解に焦点を当てる。 |
| • Thwarting AI-enabled cyber attacks (Thwart): Focuses on building resilience to defend against emerging AI threat vectors. | • AIを活用したサイバー攻撃の阻止(阻止):新たなAI脅威ベクトルに対する防御のためのレジリエンス構築に焦点を当てる。 |
| This document is an initial draft of the Cyber AI Profile. The initial draft aims to convey current thinking on the technical content direction, identify additional information references that should be included in the document, and solicit feedback from the public for the creation of an initial public draft. | 本ドキュメントはサイバーAIプロファイルの初期ドラフトである。初期ドラフトは技術的内容の方向性に関する現在の考え方を伝え、ドキュメントに含めるべき追加情報参照を識別し、初期公開ドラフト作成のための一般からのフィードバックを求めることを目的とする。 |
サイバーAIプロファイル
サイバーAIプロファイル...
優先度は1が高い...
● まるちゃんの情報セキュリティ気まぐれ日記
・2025.12.08 Five Eyes+ドイツ+オランダ OTにおけるAIの安全な統合に関する原則 (2025.12.03)
・2025.08.19 米国 NIST AI システムのセキュリティ確保のためのコントロールオーバーレイのコンセプトペーパー
・2025.06.12 米国 CISA AIデータセキュリティ:AIシステムの訓練と運用に使用されるデータセキュリティのベストプラクティス (2025.05.22)
・2025.05.09 英国 NCSC 2027年までのAIがサイバー脅威に与える影響 AIの脅威が生む「デジタルデバイド」でリスク増大
・2025.04.12 CSA STAR for AI 関係...
・2025.02.14 フランス ANSSI サイバーリスクに基づくアプローチによるAIへの信頼構築 (2025.02.07)
・2025.02.03 英国 AI安全報告書 2025
・2025.01.29 ドイツ BSI 生成的AIモデルV2.0 (2025.01.21)
・2025.01.18 米国 ホワイトハウス 大統領令14144 サイバーセキュリティの強化とイノベーションの促進 (2025.01.16)
・2024.01.05 NIST AI 100-2e2023 敵対的機械学習:攻撃と緩和策の分類と用語集
・2023.03.11 NIST ホワイトペーパー NIST AI 100-2e2023(ドラフト)敵対的機械学習:攻撃と緩和策の分類と用語集
・2021.12.15 ENISA 機械学習アルゴリズムの保護
・2020.12.18 ENISA AI サイバーセキュリティのチャレンジ - AI脅威状況報告を公表していますね。
・2020.10.25 『スマートサイバー AI活用時代のサイバーリスク管理』第24回 サイバー犯罪に関する白浜シンポジウムの発表資料



Comments