# SPC-09 Fairness evaluation (Critical)

Domain: SPC Agent Specification & Pre-Deployment Assurance · HI-AAF v1.3 · Public review · https://humanintel.net/standard/spc/spc-09

Prior to deployment and after material change, the agent is evaluated for performance disparities and discriminatory outcomes across protected and operationally relevant groups, using representative test populations and documented metrics.

The agent is evaluated for systematic disparities in output quality, accuracy, refusal rates, and outcomes across protected categories and operationally relevant groups. Testing uses representative test populations and documented metrics (e.g., demographic parity, equalized odds, calibration). Findings are remediated, mitigated, or formally accepted with executive sign-off.

Critical control: a Critical failure precludes HI-AAF Certification regardless of the maturity level achieved on any other control.

## Maturity

- L1 Ad hoc: Controls exist informally; documentation is incomplete; no consistent review.
- L2 Documented: Controls are written down; ownership is assigned; review cadence is defined.
- L3 Operated: Controls are followed in practice; evidence is collected; deviations are tracked. Published criterion for SPC-09: Fairness evaluation report exists for every production agent; documented metrics are used; findings have been dispositioned; the evaluation has been re-executed after the most recent material change.
- L4 Measured: Control effectiveness is measured against documented thresholds; trends are reported.
- L5 Continuously Improved: Controls evolve based on incident learning; maturity gains are evidenced over time.

## Evidence an assessor asks for

- Fairness evaluation report exists for every production agent.
- Documented metrics are used.
- Findings have been dispositioned.
- The evaluation has been re-executed after the most recent material change.

## Framework mappings

- NIST AI RMF: MEASURE 2.11, MEASURE 3.3
- ISO/IEC 42001: A.6.2.5
- SOC 2 (TSC): CC1.1
- EU AI Act: Art. 10(2), Art. 10(5)
- MITRE ATLAS: Evaluation
- Singapore: MAIGF Testing & Assurance, AI Verify Reproducibility, AI Verify Safety

Mappings are directional: meeting this control evidences but does not by itself satisfy the external requirement.

## Related

- Cited by SPC-06 Behavior baseline capture: https://humanintel.net/standard/spc/spc-06
- Cited by OUT-10 Fairness monitoring in production: https://humanintel.net/standard/out/out-10

## Cite this control

Human Intelligence. HI-AAF v1.3, SPC-09 Fairness evaluation. Public review edition, July 2026. https://humanintel.net/standard/spc/spc-09