The Human Layer Audit: Measuring Accountability in AI Systems

Paper 3 in The Human Layer series. This paper provides a scoring and assessment framework for measuring whether human oversight in AI systems is structurally sound. It introduces a risk-tiered maturity model (five components, five levels, three risk tiers), identity anchoring requirements, compliance traceability specifications, escalation threshold calibration, and a floor-based scoring methodology designed to prevent organizations from averaging away architectural weaknesses. The framework addresses adversarial resilience through verification sampling with probe safety constraints, dual-track calibration for delayed ground truth environments, and telemetry governance principles. It includes a self-assessment instrument for immediate operational use. Paper 1 (DOI: 10.5281/zenodo.19119699) established the economic and institutional case. Paper 2 (DOI: 10.5281/zenodo.19120077) specified the architecture. This paper operationalizes both into a measurable, auditable standard.