Evidence layer
Probative surfaces brought into scope by this page
This page does more than point to governance files. It is also anchored to surfaces that make observation, traceability, fidelity, and audit more reconstructible. Their order below makes the minimal evidence chain explicit.
- 01Canon and scopeDefinitions canon
- 02Response authorizationQ-Layer: response legitimacy
- 03AttestationQ-Attest protocol
- 04Audit reportIIP report schema
Definitions canon
/canon.md
Opposable base for identity, scope, roles, and negations that must survive synthesis.
- Makes provable
- The reference corpus against which fidelity can be evaluated.
- Does not prove
- Neither that a system already consults it nor that an observed response stays faithful to it.
- Use when
- Before any observation, test, audit, or correction.
Q-Layer: response legitimacy
/response-legitimacy.md
Surface that explains when to answer, when to suspend, and when to switch to legitimate non-response.
- Makes provable
- The legitimacy regime to apply before treating an output as receivable.
- Does not prove
- Neither that a given response actually followed this regime nor that an agent applied it at runtime.
- Use when
- When a page deals with authority, non-response, execution, or restraint.
Q-Attest protocol
/.well-known/q-attest-protocol.md
Optional specification that cleanly separates inferred sessions from validated attestations.
- Makes provable
- The minimal frame required to elevate an observation toward a verifiable attestation.
- Does not prove
- Neither that an attestation endpoint exists nor that an attestation has already been received.
- Use when
- When a page deals with strong proof, operational validation, or separation between evidence levels.
IIP report schema
/iip-report.schema.json
Public interface for an interpretation integrity report: scope, metrics, and drift taxonomy.
- Makes provable
- The minimal shape of a reconstructible and comparable audit report.
- Does not prove
- Neither private weights, internal heuristics, nor the success of a concrete audit.
- Use when
- When a page discusses audit, probative deliverables, or opposable reports.
Complementary probative surfaces (1)
These artifacts extend the main chain. They help qualify an audit, an evidence level, a citation, or a version trajectory.
Citations
/citations.md
Minimal external reference surface used to contextualize some concepts without delegating canonical authority to them.
AI can cite a source… and still distort it. The presence of a link or a reference does not guarantee that the synthesis respects the perimeter, the conditions, or the limits of the original source. That is precisely the challenge addressed by proof of fidelity.
Operational definition
Proof of fidelity: a demonstration that the response produced strictly respects the content, the perimeter, and the declared limits of the canonical source, without abusive extrapolation or external fusion.
Citation vs fidelity
- Citation: mentioning a source.
- Fidelity: an exact correspondence between the statement reproduced and the statement authorized by the source.
A citation can conceal:
- an unauthorized extrapolation,
- an abusive generalization,
- a fusion with other sources,
- the removal of conditions or limits.
Common forms of distortion
1) Omission of perimeter
The response removes the conditions that frame the source: date, region, product, or version.
2) Implicit requalification
The source describes; the response prescribes.
3) Semantic fusion
The response combines several sources under cover of a single citation.
4) Normative smoothing
Nuance disappears in favor of an averaged version that appears more “acceptable.”
Quick diagnosis
- Compare sentence by sentence: is the statement reproduced identical in meaning?
- Check the perimeter: does the response preserve the limits?
- Identify additions: does the response introduce an inference absent from the source?
- Test reproducibility: is fidelity stable across multiple prompts?
How to structure enforceable proof
1) Define an explicit canon
- Clear pivot pages.
- Structured declarations.
2) Frame response conditions
- Authorized.
- Conditional.
- Prohibited.
3) Version and date
- Make the chronology of changes visible.
4) Test the canon-output gap
- Compare the AI output with the canonical formulation.
What is strategically at stake
In an environment interpreted by AI systems, visibility is no longer enough. Fidelity becomes a condition of stable existence. A citation can reassure. Only proof of fidelity stabilizes.
Recommended links
- Definition: interpretive governance
- Authority boundary: what AI can deduce, and what it must not infer
- Interpretive smoothing: why AI standardizes your thinking
- Authority conflict: what to do when two “strong” sources oppose each other
FAQ
Is a citation sufficient as proof?
No. A citation points to a source, but it does not guarantee fidelity of interpretation.
Why is distortion difficult to detect?
Because the response remains plausible and coherent, even when it exceeds the authorized perimeter.
How can fidelity be measured?
By comparing the output with a canon that is structured, versioned, and explicitly bounded.
Minimal comparison protocol
To move from citation to proof of fidelity, a minimal protocol can compare five elements:
- the object: what exactly the source is about;
- the perimeter: what the source covers and what it does not authorize one to extend;
- the modality: fact, hypothesis, rule, exception, or non-response;
- the temporality: version, date, context, validity;
- the restitution: what the output preserves, omits, requalifies, or fuses.
A citation that preserves the object but loses the perimeter and the modality is not yet a proof of fidelity.
Why the protocol must remain tied to the canon
Fidelity is not an impression of resemblance. It presupposes anchoring in surfaces that clearly publish what prevails.
That is why proof of fidelity must remain connected to the Machine-first canon, the Site role, governance files, and ideally to a measure of the canon-output gap.