Skip to content
Human Intelligence seal
Human Intelligence
§ I · Domain MON · Control 12 of 12
Control
MON-12
Edition
v.1.3
Status
Public review

MON · Continuous Monitoring & Human Oversight

MON-12Model identity verification

The model serving each request is logged as a verifiable fact and checked against the certified, pinned version. A mismatch generates an alert and is handled as an incident.

The model serving a request is treated as a logged, verifiable fact rather than an assumption. Per-step logs capture the model identity and version actually serving each request, and that identity is verified against the certified, pinned version. A mismatch, including a provider-initiated fallback to a different model, generates an alert and is handled as an incident. This control converts a silent model substitution from an invisible event into a detected one, and is the runtime complement to the fallback-assurance requirement in OUT-16.

Table MON-12.1 · Maturity criteria · v.1.3

L1 to L5 · the lowest level in scope caps the agent
L1Ad hoc

Controls exist informally; documentation is incomplete; no consistent review.

L2Documented

Controls are written down; ownership is assigned; review cadence is defined.

L3Operated

Controls are followed in practice; evidence is collected; deviations are tracked.

Per-request logs capture serving model identity and version; automated comparison against the pinned version is active; at least one mismatch, in testing or production, has been detected and routed to the incident process within the documented latency.

L4Measured

Control effectiveness is measured against documented thresholds; trends are reported.

L5Continuously Improved

Controls evolve based on incident learning; maturity gains are evidenced over time.

The v1.3 edition publishes a control-specific criterion at L3 (Operated). The other levels apply the general ladder to this control; see the maturity ladder.

Evidence an assessor asks for

  1. E.1

    Per-request logs capture serving model identity and version.

  2. E.2

    Automated comparison against the pinned version is active.

  3. E.3

    At least one mismatch, in testing or production, has been detected and routed to the incident process within the documented latency.

Table MON-12.2 · Framework mappings

4 of 7 frameworks carry a control-level reference
NIST AI RMFMEASURE 2.4, MANAGE 4.1NIST AI Risk Management Framework 1.0
ISO/IEC 42001·ISO/IEC 42001:2023 · Artificial intelligence management system
OWASP LLM Top 10·OWASP Top 10 for Large Language Model Applications
SOC 2 (TSC)CC7.1, CC7.2AICPA Trust Services Criteria for SOC 2 Type II
EU AI ActArt. 14, Art. 26EU AI Act · Regulation (EU) 2024/1689
MITRE ATLAS·MITRE ATLAS · Adversarial Threat Landscape for AI Systems
SingaporeMAIGF Incident ReportingSingapore Model AI Governance Framework / AI Verify / PDPA

Mappings are directional: meeting this control evidences but does not by itself satisfy the external requirement. A middle dot means no control-level reference is published at this edition; see the crosswalk.

Related controls and terms

Cite this control

Human Intelligence. HI-AAF v1.3, MON-12 Model identity verification. Public review edition, July 2026. https://humanintel.net/standard/mon/mon-12

Edition v.1.3 · July 2026 · Public review · Markdown · PDF

https://humanintel.net/standard/mon/mon-12

Notice

HI-AAF v1.3 is an independent framework published by Human Intelligence. It is not affiliated with NIST, ISO, AICPA, or any government or accreditation body. The framework is offered openly for review during the current draft cycle.