Skip to content

Changelog

Every change to a watched standard lands here, newest first. Entries are generated from the merged pull requests that update the reference pages, so this log doubles as the library's audit trail.

Severity follows the project rubric: version_bump (new edition/revision), substantive (changed definition, new control, changed threshold), and editorial (normally suppressed; surfaced only when notable).

Completed tooltip coverage on the remaining pages. ISO/IEC 42001 (Clauses 4–10, Annex A–D) and ISO/IEC 23894 (Clauses 4–6, Annex A–C) now carry per-page hover definitions (kept page-local to avoid the "Annex B" cross-standard collision). Every linked Key Section entry — EU AI Act articles, DORA articles, UK principles, Berkeley CLTC papers and NIST ITL pages — now shows an explanatory tooltip on mouseover of the link itself, in addition to the existing term glossary.

Added hover/focus tooltips for ~27 key terms (defined once in a shared glossary, shown wherever a term appears) and expanded each definition. Linked the Key Sections, Audit-Relevant Anchors and Cross-Framework Mapping entries to their direct sources: EU AI Act articles/annexes (artificialintelligenceact.eu), DORA articles, NIST functions (AIRC Playbook), UK principles (gov.uk) and the CSA/Berkeley pages. ISO clauses are paywalled, so those cells point to the ISO catalogue rather than fabricated deep links.

2026-06-16 — Deeper own-words explainers

Expanded the five lighter pages — ISO/IEC 42001, ISO/IEC 23894, CSA Agentic Profile, Berkeley CLTC and NIST ITL Landscape — with an "In depth (in our own words)" section. These explain the concepts in our own words (no verbatim text), so the paywalled/licence-restricted standards gain depth without reproducing any provider wording.

2026-06-16 — New pages + SEO/GEO

Added three pages: Standards at a glance (comparison of all nine), a consolidated Glossary, and an applied AI audit for enterprise CMS guide. SEO/GEO: robots.txt (search + AI crawlers, sitemap) and JSON-LD structured data (Person · WebSite · ProfessionalService) on the home page.

2026-06-16 — Fact-check pass & corrections

Intensive accuracy review of all nine pages (dates, versions, article/clause numbers, verbatim quotes, licences), cross-checked via web sources and an independent verifier. Corrections made: ISO/IEC 42006 is published (42006:2025), not a draft; the CLTC "Intolerable Risk Threshold Recommendations" is February 2025, not January 2025; the CSA Agentic Profile draft date softened to "early 2026" (public page dated April 2026). All four verbatim "Source text" quotes (NIST, EU AI Act, DORA, UK) confirmed against their primary sources.

2026-06-16 — Plain-language explainer + licence layer

Each page gained an "In plain language" explainer (our own words, not the normative text), a Key terms glossary, and a "Provenance & licence" box. Per-source license + reproduction tiers were added to the schema and seed (public-domain / open-attribution / share-alike-noncommercial / facts-and-quotation / proprietary-metadata-only) so reproduction stays within each provider's licence. No change to any underlying standard.

2026-06-16 — Library depth enrichment

All nine reference pages gained an Auditor Checklist (evidence-oriented checks) and a Cross-Framework Mapping (indicative cross-references to NIST / ISO / EU AI Act / OWASP). No change to any underlying standard — an editorial depth pass on the baseline. Future entries here track standard changes only.

2026-06-16 — Initial baseline

First baseline snapshot of all five watched standards. Each reference page was hand-authored from current primary sources and the local research library, pending the first automated routine run.

Standard Severity Note
NIST AI RMF baseline AI RMF 1.0, GenAI Profile 600-1, Adversarial ML 100-2e2025. Agentic Profile noted as forthcoming (~Q4 2026).
EU AI Act baseline Base regulation + 2026-05-07 Digital Omnibus provisional agreement (high-risk deferral to 2027/2028), pending OJ publication.
ISO/IEC 42001 baseline 2023 1st edition current; ISO/IEC 42006:2025 (cert-body requirements) at DIS stage.
CSA Agentic Profile baseline Lab Space draft v1 (2026-03-27) + CSA 2026 catastrophic-risk / AICM / STAR-for-AI programme.
Berkeley CLTC baseline Cyber-threat thresholds white paper (2026-01-22) + intolerable-risk recommendations (2025-01).
ISO/IEC 23894 baseline 2023 1st edition; AI risk-management guidance, companion to ISO/IEC 42001 and ISO 31000.
UK AI White Paper baseline 2023 pro-innovation White Paper (CP 815) + 2024 response; AI Growth Lab consultation (Oct 2025 → Jan 2026).
DORA baseline Reg. (EU) 2022/2554, applies since 2025-01-17; five pillars + ICT third-party oversight (AI vendor resilience).
NIST ITL Standards Landscape baseline ITL AI program landscape, anchored to the 2026-03-06 AI Standards Landscape webinar.