K05 assurance architecture
Autonomous defense is bounded, causal, and reversible
Direct answer
Autonomous defenders act only through pre-authorized internal actions whose physical consequence, confidence threshold, rollback, evidence, and safe-state behavior have been tested.
Evidence boundary. These records describe architecture, control relationships, test requirements, and legal-source status. They do not certify a facility, authorize a mission, prove deployment, or replace current facility-specific engineering and legal review.
Claim
Autonomous defenders act only through pre-authorized internal actions whose physical consequence, confidence threshold, rollback, evidence, and safe-state behavior have been tested.
Argument
Machine-speed containment can reduce loss only when false-positive effects cannot exceed the threat being controlled.
Evidence requirements
- allowed-action policy
- causal dependency model
- threshold and uncertainty records
- rollback tests
- false-positive consequence analysis
- model and data provenance
- decision receipts
Assurance defeaters
- unbounded tool access
- unknown physical consequence
- model drift
- poisoned telemetry
- rollback failure
- conflicting agent state
Hazards
- self-inflicted outage
- unsafe OT isolation
- cascading agent failure
- evidence loss
Recovery objectives
- stop action on uncertainty
- restore last known-good state
- retain incident chronology
- escalate unresolved state
Site-tailoring questions
- action tier
- protected process
- latency requirement
- operator authority
- maintenance state
Sources and record
NIST SP 800-82 Revision 3 — Guide to Operational Technology Security · NIST SP 800-207 — Zero Trust Architecture · White House Launches GOLD EAGLE Initiative for Cybersecurity Vulnerability Coordination