# Human-in-the-Loop Decision Matrix

Choose where AI may act, where it should recommend, and where humans must retain final authority.

## Best for
- AI governance
- Operations
- Risk teams

## Decision class
- [ ] Low consequence
- [ ] Moderate consequence
- [ ] High consequence
- [ ] Regulated/consequential

## AI role
- [ ] Automate
- [ ] Recommend
- [ ] Summarize
- [ ] Route only

## Human control
- [ ] Approval required
- [ ] Review sampling
- [ ] Escalation owner
- [ ] Override mechanism

## Evidence
- [ ] Inputs retained
- [ ] Decision reason
- [ ] Tool actions
- [ ] Outcome log

## Related UbiGrowth resources
- https://www.ubigrowth.ai/resources/ai-agents
- https://www.ubigrowth.ai/resources/ai-workflow-automation

Use this template as a planning aid. Adapt approvals, controls, and review requirements to the consequences of the workflow.