Priority 01
Human–AI work design & governance
Human–AI Collaboration Maturity · weight 25%
40.0 → 70
Rebuild the layer
+7.5 index points recoverable
Design roles around AI-augmented work so accountability strengthens as automation increases.
Current state
AI adoption is tool-led and uneven, oversight is informal, and accountability blurs where automated output enters a decision.
Target state
Redesigned role boundaries with explicit human-in-the-loop checkpoints, AI fluency baselines per role, and auditable oversight of automated decisions.
Structural shifts
- Role-by-role work redesign identifying which judgement stays human
- Documented human-in-the-loop control points on every customer- or risk-facing workflow
- AI fluency baseline and certification tied to role level, not volunteer interest
- Governed tooling catalogue replacing unmanaged local adoption
- Decision rights
- Process owners approve automation boundaries; risk and legal approve oversight controls.
- Operating cadence
- Fortnightly adoption and oversight review; quarterly control audit.
- Accountable
- Chief Operating Officer with the AI Governance Forum
- Instrumentation
- Fluency coverage by role level, share of workflows with defined oversight, exception and override rates.