Glossary
The vocabulary of Auditome and ASE. Learn the terms once and the rest of the system reads like a sentence.
ASE
Auditome Sovereign Engine. The runtime governance and evidence engine that evaluates AI-assisted actions before execution, returns a governed verdict, and produces signed evidence.
ASE Gate
The enforcement point where AI intent becomes action. It checks trace, policy, authority, and evidence before an action can proceed, and fails closed.
Auditome
The platform/company that builds execution integrity infrastructure for regulated and high-trust AI systems — controlling and evidencing what AI-assisted systems are allowed to do before they create real-world consequence.
Authority
Verified permission — of a human or system — to take an action at decision time. No authority, no action.
Drift Detection
Layer 7: four detectors (resource scope, actor context, timing, payload change) that check a later action against the certified baseline. When drift is detected, it triggers a suspension event requiring restart-time authority revalidation.
Evidence
Proof required to support a verdict or a claim. No evidence, no claim.
Execution Integrity
The discipline of controlling and evidencing AI-assisted actions at the moment intent becomes consequence.
Fail-Closed
When required proof is absent, the default is to refuse or escalate — not to let the action through.
Gate Evaluation
Layer 1 of the ASE stack: the action is checked before it becomes real.
Governance
The system can determine whether an action was permitted and enforce that decision before the action happens. Governance lives in the execution path.
Human Authority Layer
Layer 5: ASE verifies the reviewer held current authority before their review decision is allowed to stand.
Case File Assembly
Layer 8: a read-only endpoint that reconstructs the full governed history of one decision — approval, drift, suspension, redress, human review — in chronological order, with explicit gap reporting for pre-migration events.
Policy
The rule that applies to an action and what it requires before the action may proceed.
Proof Surface
The set of evidence — trace, policy, authority, receipt — that a governed decision exposes for review and reconstruction.
Reconstruction
Rebuilding exactly why an action was allowed or refused, from its trace, policy, authority, and receipt.
Redress
The path by which a decision can be challenged after the fact and re-examined under governance.
Signed Decision Receipt
Layer 2: cryptographic, tamper-evident evidence that a governed decision occurred before execution. Not an audit log.
Suspension & Recovery
Layer 6: atomic fail-closed suspension with signed receipts. Authority is revalidated at restart time, not carried forward from the original grant.
Trace
A clear record of what was requested and the context around it. No trace, no trust.
Verdict
The governed decision ASE returns: approved, rejected, or requires human review.
Verdict Coverage
Layer 4: every outcome — approval, rejection, and human-review escalation — produces signed evidence.
Visibility
Being able to see what happened: logs, dashboards, monitoring, audit trails. It does not stop unauthorized action.