InfraGuard AI is a defensive simulation. It contains no real critical-infrastructure credentials, PLC addresses, exploit code, scanning logic, or autonomous interfaces to physical systems.

All assets, telemetry, actions, model artifacts, and adversarial scenarios are synthetic. Any real-world adaptation would require explicit authorization, network isolation, safety-engineering review, human approval for consequential actions, rollback/degraded-safe modes, and validation against the operator's actual IT/OT/ICS architecture.