The AI Tutor will answer grounded in this lesson's cited source (jira_story: GOV-312) and key concepts — authority Informational. Not yet enabled.
Lesson 3 of 3 — Offline verifier tooling
Cryptographic Trust & Anchoring · Source: Jira Story GOV-312 (5 pts) · Authority: Informational (the Jira story is authoritative).
Purpose
Let anyone verify an evidence bundle without the platform.
Why it matters
Independent, offline verification is the ultimate trust guarantee.
Where this lives
Runs in the Customer AI Governance platform (components: shared). A standalone verifier checks signatures and anchors in an evidence bundle offline. Learn teaches this read-only.
Workflow
- Given an evidence bundle
- When a third party checks it
- Then they can verify it fully offline
Worked example
An auditor wants to verify an exported bundle on an air-gapped machine. Can they?
Yes — the offline verifier checks signatures and anchors with no call back to the platform.
Acceptance criteria (authoritative — GOV-312)
Given an evidence bundle, when a third party checks it, then they can verify it fully offline.
Common pitfalls
- Verification that requires calling our service, defeating independence.
Knowledge check
- Scenario: An auditor wants to verify an exported bundle on an air-gapped machine. Can they?
- Look for: Yes — the offline verifier checks signatures and anchors with no call back to the platform.
Key concepts
- offline verification
- third-party verifier
- evidence bundle
For the AI Tutor
- Summary: Let anyone verify an evidence bundle without the platform.
- When to use: teaching or answering questions about "offline verifier tooling".
- When NOT to use: producing the regulator export (that is GOV-144).
- Key concepts: offline verification, third-party verifier, evidence bundle.
Completion criteria
Learner passes the scenario knowledge check and can point to the system state that satisfies GOV-312's acceptance criterion.
Related
- Module: LEARN-MOD-010 (Cryptographic Trust & Anchoring)
- Provenance: LEARN-MAP-001 · Generation standard: LEARN-STD-002
- Next: (module complete)