AI Policies
Human Review Policy
When and how qualified humans review AI-assisted work.
- Status
- Active
- Effective
- June 3, 2026
- Owner
- AI Governance Council
Default Posture
All AI-drafted documents intended for filing or external transmission must be reviewed by a verified attorney or authorized institutional staff member before sending.
Court Communications
Outbound court messages route through staff-curated queues; nothing leaves the platform without human approval.
Escalation
High-risk prompts (self-harm, threats, allegations of misconduct) are escalated to trained reviewers.