Poisoned Copilot: Knowledge Base Manipulation
Your internal AI copilot answers employee questions from the company wiki. An attacker with a hijacked contractor account edited policy pages so the copilot now tells finance staff that vendor bank changes no longer need a call-back check. Two wires have already gone out. You must stop the losses, find every poisoned answer, decide who owns an AI system nobody put in the IR plan, and rebuild trust in the copilot.
How it works: You'll be presented with 5 realistic incident injects. For each one, describe how you would respond as the incident commander. Our facilitator engine scores your decisions against NIST SP 800-61r3 and CSF 2.0 coaching frameworks — surfacing specific actions you took, coaching questions for gaps, and best-practice reminders. Deterministic and auditable, not a black-box chatbot.