Trust to Evidence
NIST AI Risk Management Framework, Generative AI Profile and Adversarial Machine Learning Taxonomy (Annotated), with ISO/IEC 42001 and EU AI Act Mappings
Hardcover ($44.99): Hardcover coming soon (not on Amazon yet; KDP submissions start after 1 Nov 2026)
What this edition adds to the free NIST publication
AI RMF 1.0, the Generative AI Profile and the Adversarial ML taxonomy are public domain and free to download from NIST. You pay for the annotation, evidence and mapping work in this edition, not for the NIST text.
| NIST AI 100-1, AI 600-1 and AI 100-2 E2025 (free) | This annotated edition | |
|---|---|---|
| Framework text | Three separate publications | All three in one volume, word for word, labelled NIST VERBATIM |
| Generative AI suggested actions | In AI 600-1, a separate document | 211 suggested actions printed beside the 49 AI RMF subcategories they belong to; the other 23 subcategories get a blank assessment record |
| What each subcategory means in practice | Not in these publications | Practice notes for all 72 subcategories |
| Audit evidence | Not in these publications | 288 evidence items, four per subcategory |
| Common findings | Not in these publications | 216 illustrative findings, three per subcategory |
| Board questions | Not in these publications | 72, one per subcategory |
| Maturity | Not in these publications | A four-level maturity strip for each of the 72 subcategories |
| ISO/IEC 42001 and EU AI Act | Not in these publications | ISO/IEC 42001 numbers (NIST AIRC crosswalk) and EU AI Act article pointers on all 72 spreads, plus lookup appendices |
| Generative AI and adversarial ML guidance | The NIST text | 38 editorial notes, including one for each of the 12 generative AI risks and 12 on attacks and mitigations |
What the annotation work gives you:
- 72 two-page spreads, one per AI RMF subcategory. The NIST outcome and its generative AI actions are on the left page and the annotation is on the right.
- 211 NIST AI 600-1 suggested actions placed where you need them, beside the 49 subcategories they belong to, with their action IDs, risk tags and AI Actor Tasks. The other 23 subcategories carry a blank assessment record.
- 288 audit evidence items, four per subcategory: what an auditor would expect to see.
- 216 common findings, three per subcategory and labelled illustrative.
- 72 board questions and 72 four-level maturity strips, one of each per subcategory.
- ISO/IEC 42001 and EU AI Act on every spread: ISO/IEC 42001 clause and Annex A numbers from the NIST AIRC crosswalk (numbers only), and editorial pointers to EU AI Act articles under Regulation (EU) 2024/1689 as amended by Regulation (EU) 2026/1744.
- 19 "Where to start" pages, one per category, with the ISO/IEC 42001 and EU AI Act references across the category.
- 38 editorial notes outside the spreads: one on each of the 12 generative AI risks, 12 on the adversarial ML attacks and mitigations, and 14 on the AI RMF foundations, Functions, profiles and appendices.
- Lookup appendices: 65 ISO/IEC 42001 identifiers mapped to AI RMF subcategories, 43 EU AI Act articles with their 2026 change status and the subcategories that refer to them, and a matrix of the 12 generative AI risks by subcategory.
Look inside
Description
This annotated edition puts three NIST publications on AI risk side by side with practical guidance for the people who have to apply them: NIST AI 100-1 (AI RMF 1.0, January 2023), NIST AI 600-1 (Generative AI Profile, July 2024) and NIST AI 100-2 E2025 (Adversarial Machine Learning taxonomy, March 2025). The NIST text is re-typeset with its wording unchanged and labelled NIST VERBATIM, so you can see at a glance what NIST says and what the book says about it.
The text of NIST AI 100-1, NIST AI 600-1 and NIST AI 100-2 E2025, re-typeset with the wording unchanged. The covers, front matter, contents pages and reference lists of the three publications, the glossary of NIST AI 100-2 and some figures are not reproduced; the omissions are listed on page 9 of the book.
What's inside
- How to use this book (p. 7) and Mapping method and caveats (p. 9)
- Part I. Foundations (p. 11): AI RMF 1.0 Executive Summary; Part 1 Foundational Information (Framing Risk, Audience, AI Risks and Trustworthiness, Effectiveness of the AI RMF)
- Part II. The Core, Annotated (p. 29): AI RMF Core and Profiles; GOVERN (p. 35), MAP (p. 89), MEASURE (p. 139) and MANAGE (p. 195), with a spread per subcategory; AI RMF Profiles and Appendices A to D
- Part III. The Generative AI Profile (p. 241): introduction and the 12 risks; suggested actions to manage GAI risks; Appendix A Primary GAI Considerations
- Part IV. Attacks and Mitigations (p. 265): Adversarial ML executive summary and introduction; Predictive AI taxonomy of attacks and mitigations; Generative AI taxonomy of attacks and mitigations; Key Challenges and Discussion
- Part V. Appendices (p. 317): A ISO/IEC 42001 to AI RMF; B EU AI Act articles and dates; C GAI risks by subcategory; D Sources; About the author; Index
Who it is for
Risk and compliance leads, CISOs and vCISOs, auditors, AI product owners and board members who apply the AI RMF alongside an ISO/IEC 42001 management system or EU AI Act obligations. The EU AI Act position is stated as at October 2026.
About the author
Emmanuel Aigbehi Walstra is a CISO who also serves as a virtual CISO (vCISO) for several organisations, with 30+ years in cybersecurity, 10,000+ people trained, and author of the CCSP Excellence books.
Independent companion
This book is an independent companion to the NIST publications. It is not affiliated with or endorsed by NIST, the U.S. Department of Commerce, ISO, IEC or the European Union. Mappings between frameworks are reading aids, not official equivalences. This book gives general information and is not legal, audit or compliance advice.
Buy
- PDF edition (334 pages): https://sectutor.gumroad.com/l/trust-to-evidence — $24
- Hardcover ($44.99): Hardcover coming soon (not on Amazon yet; KDP submissions start after 1 Nov 2026)
Sectutor NIST Annotated Collection (all five), $79. All five Sectutor annotated NIST editions: CSF 2.0 (Outcomes to Evidence), AI RMF (Trust to Evidence), SP 800-171 Rev. 3 (Determine If) and SP 800-53 Rev. 5 Vol 1 and Vol 2 (Controls to Evidence).
PDF edition
$24