Automated Responses
Configure automated response actions for security incidents, compliance violations, and operational events. Build custom playbooks with triggers, conditions, and multi-step action sequences that execute without manual intervention.
Creating Custom Playbooks
Build your own automation workflows in four steps:
1. Define Trigger
Select what event initiates the playbook:
- Security alert — Defender, Identity Protection, or Sentinel alerts
- Compliance drift — Intune non-compliance, Trust Center policy violations
- Scheduled — Daily, weekly, or custom cron schedule
- Manual — On-demand execution from the dashboard
- Webhook — External system triggers via HTTP callback
2. Set Conditions
Narrow when the playbook should fire:
- Alert severity — High, Medium, Low
- User type — Admin, Guest, Member
- Device platform — Windows, macOS, iOS, Android
- Tenant/customer selection — Specific tenants or all managed tenants
3. Configure Actions
Define the response actions to execute:
User Actions
- Block sign-in immediately
- Revoke all active sessions
- Reset password and require MFA re-registration
- Disable account
Device Actions
- Force device sync
- Remote lock device
- Wipe device (factory reset)
- Retire device (remove corporate data)
Notification Actions
- Send email alerts to security team
- Post to Teams channel
- Send Slack notification
- Fire webhook to external system
Ticket Actions
- Create ticket in PSA system
- Assign to specific team or technician
- Set priority based on alert severity
Wait Actions
- Delay next action by specified time
- Wait for condition (e.g., device still non-compliant after 24 hours)
4. Test and Deploy
- Run in simulation mode first to review what actions would be taken
- Review simulation results and adjust conditions or actions
- Enable for production with approval gates if needed
Built-in Response Playbooks
| Playbook | Trigger | Key Actions | Status |
|---|---|---|---|
| Compromised User Response | High-risk sign-in | Block, revoke sessions, reset password, notify SOC | Active |
| Device Non-Compliance | Intune non-compliance | Notify owner, force sync, block access after 24h | Active |
| Suspicious Mail Rule | External forwarding rule | Disable rule, alert security, notify user | Active |
| MFA Not Registered | Registration deadline passed | Send reminders, block after 14 days | Paused |
| Stale Account Cleanup | No sign-in for 90+ days | Notify manager, disable, remove licenses | Active |
Approval Gates
For sensitive automated responses, require human approval before execution:
- Pause playbook and send approval request via email, Teams, or dashboard
- Set timeout for auto-rejection (default 4 hours) or auto-approval
- Require multiple approvers for critical actions (e.g., account wipe)
- Approval decisions are logged in the audit trail
Response Chaining
Chain multiple responses with conditional logic:
- Initial action — Low-impact response (send notification)
- Wait period — Configurable delay (4 hours, 24 hours, 7 days)
- Condition check — Evaluate if issue persists
- Escalation action — Higher-impact response (block access, create ticket)
- Final action — Destructive response if needed (wipe, disable)
Best Practices
- Start with simulation mode — Run playbooks in simulation to review what actions would be taken before enabling live execution
- Use approval gates for destructive actions — Require human confirmation for account disabling, data wiping, or license removal
- Layer your responses — Start with low-impact actions (notifications) and escalate to higher-impact actions (blocking) over time
- Monitor success rates — Track which automated responses succeed and which fail to continually improve reliability
Warning: Destructive actions like wipe and disable cannot be easily reversed. Always use approval gates for high-impact automated responses.
API Reference
GET /api/automation/playbooks— List all playbooksPOST /api/automation/playbooks— Create new playbookPUT /api/automation/playbooks/:id— Update playbook configurationPOST /api/automation/playbooks/:id/run— Manually trigger playbookPOST /api/automation/playbooks/:id/simulate— Run in simulation modePOST /api/automation/approvals/:id— Approve or reject pending action
Last updated on