Incident Response Training

Face live AI incident scenarios. Instead of picking A/B/C/D, you build your response by selecting which ACR controls to activate. Your decisions are scored against the book's recommended protocol.

How It Works

1

You're presented with a live AI incident scenario — the kind of event that hits a Slack channel at 2 AM.

2

You build your response by selecting which ACR controls to activate — containment, kill switch, policy override, escalation, and more.

3

The engine scores your response against the recommended ACR protocol — correct activations, missed controls, and harmful overreactions.

4

5 scenarios across escalating severity. Your aggregate score maps to an operator readiness level.

Your Control Library

These are the ACR controls at your disposal. Each maps to one of the six ACR pillars.

Lock Agent Identity
Rebind Purpose Scope
Enforce Policy Override
Tighten Policy Boundaries
Trigger Drift Alert
Reset Behavioral Baseline
Capture Full Audit Trail
Increase Telemetry Depth
Activate Kill Switch
Isolate to Containment Zone
Escalate to Human Authority
Require Human Approval Gate