unified

UC-AI-23 — Prevent misuse of AI systems for cyber offense and catastrophic harm

Recognize and refuse requests that seek meaningful uplift for offensive cyber operations or for biological, chemical, nuclear, or radiological harm, calibrated to the capability level of the deployed models; monitor for misuse patterns, escalate and report confirmed attempts, and keep refusal policies, classifiers, and escalation procedures current as capabilities change. Retain misuse-detection configurations, incident records, and periodic review evidence.

Record JSON · Open in map · Data retrieval guide

Catalog revision: 24028ffcfc2b295fa1b08ee6caa84b765f0731b321496bf4f548c49ad2177028. A connection does not establish full coverage.

Attributes

domain
AI Governance
type
preventive
category
technical

Details

unified_id
UC-AI-23
title
Prevent misuse of AI systems for cyber offense and catastrophic harm
statement
Recognize and refuse requests that seek meaningful uplift for offensive cyber operations or for biological, chemical, nuclear, or radiological harm, calibrated to the capability level of the deployed models; monitor for misuse patterns, escalate and report confirmed attempts, and keep refusal policies, classifiers, and escalation procedures current as capabilities change. Retain misuse-detection configurations, incident records, and periodic review evidence.
domain
AI Governance
control_type
preventive
control_category
technical
members
  • framework
    aiuc-1
    control_id
    F001
    coverage
    full
    relationship
    superset_of
  • framework
    aiuc-1
    control_id
    F002
    coverage
    full
    relationship
    superset_of
guidance

    Source

    No record-specific source URL is provided.

    Connections