Multi-agent isolation
Objective
Establish a repeatable and accountable capability for multi-agent isolation during AI security incident handling.
Control / requirement
The organization should define ownership, decision criteria, technical procedures, dependencies, and escalation conditions for multi-agent isolation.
Business impact
Failure may increase incident duration, uncertainty, evidence loss, propagation, or regulatory exposure.
Validation approach
Inspect approved procedure; test through scenario or technical exercise; verify retained evidence and action records.
Expected evidence
Policy or playbook; exercise or incident record; technical logs; decision and approval evidence.
Mapping and source
Prototype website catalogue — replace with the authoritative approved AI-IRF control record.
Implementation guidance
Translate the requirement into system-specific procedures, responsible roles, technical configurations and review conditions. Retain evidence showing both design and operating performance. Where the source provides no separate implementation instruction, do not infer that a single technical mechanism is sufficient.
Assessment considerations
- Confirm scope and applicability.
- Inspect control design and responsible ownership.
- Test representative operation and adverse conditions where appropriate.
- Evaluate evidence provenance, completeness and contradictory evidence.
- Record limitations and notably absent outcomes.