Saltar al contenido principal

Commit to an AI Incident Mini-Runbook

Commit to documenting a minimal incident runbook for one AI workflow.

ai-risk-management5 min read

AI incident mini-runbook for one workflow Contain, preserve, correct, improve A customer-facing AI answer invents policy, price, legal, refund, safety, or security guidance. An internal AI workflow exposes sensitive employee, customer, contract, code, or credential data. An AI agent takes or recommends an action outside its approved permission scope. Trigger and severity Define what counts as an AI incident for this workflow: escaped harmful output, sensitive disclosure, unsafe action, discriminatory recommendation, or threshold breach. Containment owner Name who can pause the feature, disable a tool, switch to manual review, or roll back to a known-good workflow. Evidence to preserve Save…

Read the full lesson

Sign up free: one personalized lesson every day, matched to your role and goals.

← Back to library

Contáctanos