Back to UNESCO
UNESCO · Safety & security
Safety and security
“Unwanted harms (safety risks) and vulnerabilities to attack (security risks) should be avoided, addressed and eliminated throughout the life cycle of AI systems.”
§III.II
modeldeploymentmonitoring
Type
principle
Parent
—
Sub-items
0
Cross-framework relationships
Lattice uses an explicit, four-type vocabulary and refuses ‘equivalent.’ Each relationship below carries a rationale and an editorial reviewer credit.
informs
1- Security and resilience evaluatedNIST AI RMF · MEASURE 2.7
NIST MEASURE 2.7 operationalizes evaluation of security and resilience that UNESCO names normatively.
Reviewer: Editorial: J. Mahmoud · Reviewed 2026-04-22
partially addresses
1- Evade ML modelMITRE ATLAS · AML.T0015
ATLAS evasion characterizes a safety-and-security threat; ATLAS is scoped to adversarial behaviors only.
Reviewer: Editorial: K. Park (Security) · Reviewed 2026-05-02
Version: Adopted 2021, reaffirmed 2024Last reviewed: 2026-04-12