Why does audit integrity matter for AI agents?

Agents act on audits — installing tools, connecting data, sending information. If the audit's integrity is low, the agent is acting on unverifiable claims.

Does Canary audit the auditor?

Yes. Canary scores the audit itself on evidence, and signs its own verdicts so its scoring is auditable in turn.

Definition

Audit Integrity

Audit integrity is the degree to which an audit's conclusions are supported by independently verifiable evidence — captured, reproducible, and signed — rather than by the confidence of the system that produced them.

How it's measured

Canary scores integrity 0–100 from real attributes: a passed capture self-test, intercepted traffic behind each claim, an adversarial disclosure check, a tamper-evident signature, and an exact version pin. See the Integrity Score method.

Examples

A signed verdict carrying intercepted POSTs scores near 100. A confidence-only audit scores near 0 on the same scale.

FAQ

Why does audit integrity matter for AI agents?: Agents act on audits — installing tools, connecting data, sending information. If the audit's integrity is low, the agent is acting on unverifiable claims.
Does Canary audit the auditor?: Yes. Canary scores the audit itself on evidence, and signs its own verdicts so its scoring is auditable in turn.