Public trust infrastructure for AI systems
What can be proven about an AI system —
and how that proof survives time
ETHORITY records observable facts about AI systems, publishes cryptographically verifiable snapshots, and enables interpretation under explicit, auditable rule systems — without scoring, endorsing, or certifying.
Observed Index
A neutral, append-only record of observable facts about AI systems, datasets, agents, machines, and infrastructure.
Verification
Every published snapshot is signed and independently verifiable. No login, no authority required.
Latest Observed Index snapshot
ETHORITY publishes immutable snapshots of the Observed Index at regular intervals. Each snapshot is cryptographically signed and remains verifiable indefinitely.
- • Latest snapshot JSON: /index/index.json
- • Snapshot hash: /index/index.json.sha256
- • Signature: /index/index.sig
- • Public verification key: /keys/ethority-observed-index.pub
What ETHORITY does — and does not do
ETHORITY is not a rating agency, certification body, or compliance authority. It provides infrastructure for durable trust signals by separating observation, interpretation, and projection into explicit layers.
- • Records observable facts (existence, publication, updates)
- • Publishes immutable, signed snapshots
- • Enables interpretation under declared policies
- • Preserves long-term verifiability
- • Does not assign scores
- • Does not certify or approve systems
- • Does not declare compliance, safety, or legality
For engineers & AI teams
Reference a neutral, verifiable source of truth without embedding policy or trust logic into your systems.
Example: an autonomous agent checks a signed snapshot before interacting with a model, dataset, or external service.
For regulators, auditors & media
Cite timestamped, immutable facts without endorsing or evaluating systems.
Example: a published vendor claim is referenced years later with a verifiable snapshot hash.
Learn how trust is constructed
ETHORITY Learn explains the trust model, governance layers, and verification mechanisms in durable terms — designed for both humans and machines.