ETHORITYAI Trust Infrastructure

ETHORITY Specification

Protocol 1.1 — Trust Lifecycle Standard

This specification defines how trust in AI systems is represented, transitioned, preserved, and degraded over time.

Lifecycle overview →Transition pricing →
L0 ObservedL1 DeclaredL2 VerifiedL3 AnchoredL4 Monitored
L0 Observed → L1 Declared → L2 Verified → L3 Anchored → L4 Monitored
Trust is governed as a deterministic state machine. Each transition is explicit, time-bound, and ledger-recorded.

Document status

Version: 1.1
Status: Active specification draft
Maintained by: ETHORITY

Purpose

The purpose of this protocol is to formalize trust in AI systems as a deterministic lifecycle state machine governed by explicit rules.

Trust is a public state. It can be referenced externally and verified independently.
Trust is time-bound. Snapshots reflect facts at time T, not forever.
Trust degrades without continuity. Monitoring maintains state; lapses trigger downgrade.

Scope and non-goals

This protocol applies to AI systems including models, agents, decision-support systems, and integrated deployments where drift, updates, or operational changes can alter trust over time.

ETHORITY does not:
  • Certify safety
  • Guarantee legal compliance
  • Replace regulatory authorities
  • Assign qualitative scores

Definitions

  • AI System: A defined deployable configuration producing outputs from inputs.
  • Snapshot: A deterministic record of observable facts at time T.
  • Verification: Cryptographic proof of artifact integrity.
  • Anchoring: Public timestamp preservation of a verification state.
  • Monitoring: Ongoing lifecycle continuity enforcement.
  • Downgrade: Explicit reduction of lifecycle state due to lapse or rule violation.

Open glossary →

Lifecycle states

Observed — Existence recorded.
Declared — Operator asserts identity and scope.
Verified — Integrity proven via signature validation.
Anchored — Verification timestamp preserved publicly.
Monitored — Continuity conditions actively enforced.

States are explicit. Transitions are explicit. Nothing is implied.

Transitions, continuity, downgrade

Each transition corresponds to a governance action. Monitoring preserves state continuity. If continuity conditions are not met, trust degrades.

Transitions

  • Observed → Declared: identity linkage
  • Declared → Verified: artifacts + signature validation
  • Verified → Anchored: timestamp anchoring
  • Anchored → Monitored: continuity governance activation

Downgrade triggers

  • Monitoring inactivity
  • Undeclared material system change
  • Invalidated artifacts
  • Declared policy non-compliance
Key rule: Verification is not permanent without monitoring.

Separation of institution and ledger

ETHORITY defines lifecycle governance rules. The Trust Ledger publishes immutable, signed artifacts. The institution defines; the ledger records.

Access Trust Ledger ↗  ·  Browse AI Registry →

Limitations

Protocol 1.1 structures trust continuity and verifiability. It does not replace regulatory frameworks or legal oversight.

Reading guide

This Protocol is the normative definition of the lifecycle model. The Lifecycle page is descriptive and optimized for humans.

Open Lifecycle overview →