# Sovereign AI system definition

## Decision metadata

- System ID:
- Owner:
- Version:
- Decision date:
- Review date:

## Purpose and approved outcomes

State the users, decisions, and outcomes the system is permitted to support. List prohibited uses separately.

## Actors and identities

| Actor | Identity source | Authorized actions | Approval owner |
| --- | --- | --- | --- |
| | | | |

## Assets and dependencies

| Asset | Classification | Source/provenance | Location | Lifecycle owner |
| --- | --- | --- | --- | --- |
| | | | | |

## Trust boundaries and data flows

Record each boundary crossing, protocol, identity, allowed data, enforcement point, and evidence source.

## Control and evidence backlog

| Control objective | Enforcement | Evidence | Gap | Owner | Due date |
| --- | --- | --- | --- | --- | --- |
| | | | | | |

## Assumptions, limitations, and exit triggers

Document unverified assumptions, unsupported conditions, portability constraints, and events that require reassessment.