Reference
Glossary
Stable terms used across ETHORITY surfaces. This glossary exists to prevent drift across teams, audiences, and time.
Observed
A recorded factual statement about an AI system or artifact, stored append-only and published without interpretation.
Note: Observed ≠ verified ≠ governed.
Observed Index
ETHORITY’s public, append-only snapshot of observed items and related metadata.
ETN
ETHORITY Trust Name: a stable identifier for an entity in the observed layer (e.g., etn:observed:system:<slug>).
Snapshot
A published, versioned export of observed records, intended for citation and independent verification.
Detached Signature
A signature stored separately from data (index.sig signs index.json).
Integrity
The property that published artifacts are authentic and unmodified, verifiable via public key cryptography.
Verification
The act of validating integrity (signature + key). It is not a safety, quality, or compliance claim.
Governance Layer
Interpretation of observed facts under explicit rule systems (EU AI Act, ISO, internal policies). Versioned and contestable.
TrustCycle
A governance run that produces a recorded interpretation, deltas, and artifacts derived from observed facts.