Canon term · MARQUE questions

Auditability

The property that a past decision can be reconstructed exactly — same inputs, same model state, same evidence, same output — rather than merely narrated from a log.

  • StatusIndustry term
  • AdoptionWidely used
  • LayerGovernance

Why it matters

Applies to compliance, internal audit and, on request, the examiner. It bites retrospectively and without warning, typically years later and typically about the one decision nobody expected to have to explain. It fails when any component of the decision was not captured: model version, memory state, retrieval corpus as it stood, prompt version, parameters. A trace that records the call chain but not the model state produces a narrative about a decision, not the decision. This is the best-covered MARQUE (a reference framework organized around eight questions) question in existing control frameworks — 131 of 230 objectives touch it — and still the most commonly failed in practice.

Where it comes from

MARQUE question 04. The recordkeeping tradition is decades old in financial services; its extension to model state is not.

No published instrument defines this term yet; the note above is its origin.

What it is not

These are routinely confused with Auditability. The distinctions have consequences for how a system is governed.

  • Logging

    Logging records that something happened. Auditability requires that everything needed to re-derive it was captured.

  • Explainability

    Explainability is why the model produced an output. Auditability is proof of what the output was and what produced it. A system can be fully auditable and entirely inscrutable.

How it connects

MARQUE questions it addresses

  • AuditabilityCan a decision be reconstructed months later, exactly?

Connected through the MARQUE crosswalk, an editorial mapping. These objectives from the Financial Services AI Risk Management Framework are assigned to the same questions that this term addresses. The ones sharing most of this term's vocabulary are shown first, as a reading aid. The assignment is to the question, not to the term: the source does not map this term to any objective directly.

Question: Auditability · 131 control objectives assigned

  • GV-6.1.5Third-Party AI Risk DocumentationGovern
  • MP-4.1.1Legal Risk Management ProcessesMap
  • MS-2.4.4Incident Investigation and Root Cause AnalysisMeasure
All 131 objectives assigned to Auditability

Cite and reuse

Cite the identifier QIS-TERM-00028, not the URL: identifiers are stable, URLs may change. This entry is free to read, quote and index under the dual license. Corrections sent to editor@qisfund.com are published.

Technical details: identifier, graph relationships and record
QIS-TERM-00028: outbound relationships as published in api/graph.json and api/lexicon.json
PredicateTarget identifierTargetPredicate meaning
addressesmarque:auditabilityAuditabilityThe subject speaks to the object as a question or concern.
relatedToterm:audit-trailAudit TrailAn association too weak or too general for a stronger verb.
relatedToterm:decision-recordAgent Decision RecordAn association too weak or too general for a stronger verb.
Canonical identifier
QIS-TERM-00028
Tier
Canonical industry term
Domain · layer
MARQUE questions · Governance
Semantic aliases
None recorded
First published
2026-08-02
Last reviewed
2026-08-02 · 180-day cycle