Organisations are running on AI decisions no one has audited

Decision Clarity (DC) — Are roles, responsibilities, and decision boundaries between humans and automated systems explicitly defined and understood? Covers override capability, decision ownership, escalation architecture, and explainability.

Cognitive Sustainability (CS) — Are the cognitive demands placed on operators within sustainable limits — and do people retain the practiced skill to intervene effectively when the system fails or produces unexpected output?

Psychosocial Safety (PS) — Do operators experience genuine agency, meaningful work, and the organisational safety to raise concerns and act on their judgment — without fear of blame or unreasonable surveillance? Anchored in eudaimonic work theory and the NSW WHS Act 2026.

Work Design (WD) — Has the work system been deliberately designed to preserve human authority — or has automation been layered over roles that were never redesigned to match? Links directly to the Intelligent Work Blueprint.

Each dimension is scored across governance questions weighted by consequence. Override Capability and Operational Skill Currency carry the highest weights within their dimensions — because their failure modes are the most acute and least recoverable..

• Map

              Identify all critical decision points and the current human-system authority split.

• Measure

              Score each decision point across the four HAI dimensions.

• Model

               Define the target state of authority distribution aligned to risk and strategy.

• Monitor

              Establish ongoing measurement and governance cadence.