Example: Verified Response
This page shows what a complete AIEP verified response looks like — every field explained.
The question
How can an AI answer be verified?
The response
An AI answer can be verified by binding claims to evidence artefacts,
hashing those artefacts, and validating the response against a declared
evidence rail.
This is not just text. Attached to it is an evidence rail and a validation result.
The evidence rail
Three artefacts were retrieved and hashed before this response was returned:
| Field | Artefact ev_001 |
|---|---|
| Source | aiep.dev/schema/evidence.json |
| Confidence | verified |
| Hash | f3a2c4e0b6a7d991... |
| Retrieved | 2026-04-25T00:00:00Z |
| Status | passed |
The hash is a SHA-256 fingerprint of the source content at retrieval time. If the source changes, a future replay will detect the discrepancy.
The validation result
Schema validation: passed
Hash validation: passed
Evidence rail present: passed
Replay metadata present: passed
Status: PASSED
Validation is performed against the AIEP canonical schema. A response cannot return PASSED without completing all checks.
Full JSON
The complete machine-readable version of this response:
{
"aiep_version": "2.0.0",
"question": "How can an AI answer be verified?",
"answer": "An AI answer can be verified by binding claims to evidence artefacts...",
"evidence_rail": [
{
"artefact_id": "ev_001",
"source_url": "https://aiep.dev/schema/evidence.json",
"confidence_tier": "verified",
"content_hash": "f3a2c4e0b6a7d991d3e2f5a9c8b7e6f1d0c9b8a7e6f5d4c3b2a1908172635445",
"retrieved_at": "2026-04-25T00:00:00Z",
"validation_status": "passed"
}
],
"validation": {
"schema_valid": true,
"hashes_present": true,
"evidence_rail_present": true,
"status": "passed"
},
"generated_at": "2026-04-25T00:00:00Z"
}
Try it yourself
Verification Playground → — run this flow interactively
See also: Verify AI Output · RAG vs AIEP · Example: Evidence Rail · Example: Audit Trail