Trust Lifecycle
Trust is governed as a lifecycle state machine.
Trust is not a certificate. It is a time-bound condition that transitions, persists, and degrades under defined governance rules.
Lifecycle overview
Every AI system progresses through explicit lifecycle states. These states describe the degree to which facts are recorded, verified, preserved, and continuously governed.
Movement between states represents governance actions — not subscription tiers, and not marketing labels.
State definitions
A public record that a system exists at time T. Append-only. No endorsement or interpretation.
The operator links identity and structured metadata. Assertions become visible, but integrity is not yet proven.
Submitted artifacts are cryptographically signed and independently verifiable. Integrity is provable at a specific moment.
Verification state is preserved through public timestamp anchoring. Historical rewriting becomes computationally impractical.
Ongoing lifecycle oversight is active. Drift detection and governance review operate on defined cadence.
Continuity
Verification alone reflects a historical truth. Monitoring ensures that truth remains aligned with declared conditions over time.
Degradation
Trust states are conditional. If continuity breaks or material conditions change without declaration, the system reverts to its last valid state.
Downgrades are explicit and recorded. Nothing is silently rewritten.
Institution and ledger
ETHORITY defines lifecycle states and governance logic. The Trust Ledger publishes signed, immutable records of those states.
Institution defines. Ledger records.
Binding rules
The lifecycle model is formally defined in Protocol 1.1. This page is descriptive. The Protocol document is normative.