Cross-Framework Translation

Indicative equivalences by framework

Published framework

NIST AI RMF, ISO/IEC 42001, and EU AI Act equivalences are indicative readings, not official certifications.

NIST AI RMF

  • Security → NIST Manage + Measure (misuse safeguards, testing, incident handling)
  • Stability → NIST Manage (continuity, monitoring, recovery)
  • Coherence → NIST Map + Measure (risk identification, behavior measurement)
  • Reputation → NIST Govern (external trust and accountability)
  • Transparency → NIST Govern (documentation, explainability)

Estimated NIST AI RMF maturity tier: Tier 3/4

ISO/IEC 42001

  • Security → clause 6 (risk treatment) + clause 8 (operational controls)
  • Stability → clause 8 (Operations) + clause 10 (Improvement)
  • Coherence → clause 6 (Planning) + clause 9 (Performance evaluation)
  • Reputation → clause 7 (Support)
  • Transparency → clause 4 (Context) + clause 5 (Leadership)

Estimated ISO 42001 coverage: 88%

EU AI Act

Articles 15, 16, 26, 50 mapped to Kanon dimensions.

Article 15 : Aligned

Kanon mapping: Security + Stability

Article 16 : Aligned

Kanon mapping: Security + Stability + Coherence + Transparency

Article 26 : Aligned

Kanon mapping: Security + Transparency

Article 50 : Aligned

Kanon mapping: Transparency

Sources used

Evidence preserved for this rating

Review methodology

Official site

Official reference review

Batch 5 reference reviewed for GitLab Duo Agent Platform (GitLab). Classification retained: Orchestrator Agent.

Open source

Batch 5 scoring report

Batch 5 scoring report

Imported from batch5-agents-148-172. Final weighted score 82/100, rating A, EU AI Act status Aligned.

Open source

Evaluation notes

Batch 5 dimension summary

Security 82, Stability 84, Transparency 76, Coherence 80, Reputation 86. Security cap: non (désactivé).

Open source

Dispute

If new material evidence changes the public record, operators can request a review through the public challenge channel.

Dispute this score

This agent has no Verified badge. Claim your public verification record.

Verify this agent →