Data
“We need clean data before we can use AI.”
The assumption
AI adoption must wait until enterprise data is fully cleansed and harmonised across the organisation.
Governance Reality
This assumption applies primarily to high-impact automated decision systems.
The 10-Dan framework recognises that AI use varies by data dependency:
Automated decision systems
Require mature, auditable data foundations aligned with Dan 5+ governance standards.
Assistive AI (drafting, summarisation, process support)
Can operate under human oversight at Dan 1–3.
Analytical AI
Requires improving data discipline at Dan 3–4.
Reframe
Data maturity and AI maturity develop in parallel—not sequentially.
Dan 1–3 allows controlled, low-risk experimentation while data governance strengthens.
Risk
“AI is too risky for a Local Authority”
The assumption
AI introduces unacceptable political, legal, or reputational exposure.
Governance Reality
Unmanaged AI is risky. Governed AI can introduce structured logging, monitoring, and traceability that may be more systematic than informal human processes.
From Dan 3 onward, the framework embeds:
Monitoring and audit mechanisms
Risk classification
Human-in-the-loop controls
Risk is not ignored, it is structured.
Reframe
The risk is unmanaged deployment—not AI itself.
The 10-Dan model exists precisely to prevent reckless scaling and ensure AI use aligns with institutional risk appetite.
Control
“If we use AI, we’ll lose control.”
The assumption
AI removes human judgement and weakens accountability.
Governance Reality
Loss of control occurs when AI is adopted informally, without structure or oversight.
The 10-Dan framework embeds control progressively:
Defined governance roles at Dan 3
Deployment approval processes at Dan 4
Independent oversight at Dan 5
Reframe
Governed AI can strengthen institutional control through transparency, monitoring, and structured override mechanisms.
Maturity strenghtens accountability.
Legal
“We could be sued if AI makes a mistake.”
The assumption
AI introduces new and unacceptable legal exposure.
Governance Reality
Legal exposure arises primarily from governance failure rather than from the mere existence of the technology.
The 10-Dan framework embeds:
Board-level accountability
Ethical review gates
Formal risk classification
Auditable decision trails
Legal exposure is managed through structure.
Reframe
Accountability is determined by governance quality.
AI does not remove statutory duties—it requires them to be explicitly structured, documented, and supervised.
Organisational
“AI will replace staff.”
The assumption
AI adoption is a workforce reduction strategy.
Governance Reality
In local authorities, AI is primarily a capacity and consistency tool, not a substitute for professional judgement.
From Dan 7, the framework embeds:
A responsible override culture
Workforce AI literacy
Psychological safety to challenge AI outputs
AI maturity should strengthen professional capability rather than sideline it.
Reframe
AI maturity strengthens workforce capability; it reinforces accountability, professional judgement and institutional responsibility.
Pace
“If we don’t move fast, we’ll fall behind.”
The assumption
Speed determines competitiveness.
Governance Reality
Public institutions are accountable before they are innovative.
The 10-Dan progression deliberately ties scaling to governance integrity thresholds.
Reframe
Disciplined progression builds durable capability.
Acceleration without governance discipline increases the likelihood of audit challenge.
National Standards
“We should wait until the central government provides a national standard.”
The assumption
Responsibility can be deferred upward.
Governance Reality
AI is already being used informally in most organisations.
Dan 1 explicitly recognises ad-hoc experimentation.
Waiting does not eliminate AI use—it removes governance from it.
Reframe
Local governance maturity can develop within national regulatory principles.
Proactive discipline reduces exposure.
Accountability cannot be outsourced.
Just an IT project
“AI is just another IT project.”
The assumption
AI adoption sits solely within digital or IT teams.
Governance Reality
The 10-Dan model operates at institutional level, not technical level:
- Strategy & leadership
- Governance & risk
- Ethical review
- Workforce literacy
- Monitoring & audit
AI maturity reflects institutional maturity.
Reframe
AI scaling without board-level governance alignment is structurally unstable.
Technology implementation alone does not constitute AI maturity.
Ban it
“We should ban generative AI internally.”
The assumption
Prohibition eliminates risk.
Governance Reality
Blanket bans often create:
- Shadow AI use
- Unmonitored data exposure
- Loss of institutional oversight
Dan 1 recognises that informal use already exists.
Dan 2–3 establishes controlled pathways for safe experimentation.
Reframe
Guardrails reduce unmanaged risk.
Governed adoption is safer than unmanaged prohibition.
Transparency
“AI decisions are opaque and unexplainable.”
The assumption
All AI systems operate as black boxes.
Governance Reality
Explainability requirements scale with impact.
Low-impact assistive uses require proportional oversight.
High-impact automated decisions require robust documentation, validation, and auditability.
From Dan 8-10, the framework increases oversight and independent validation.
Reframe
Proportional governance is the answer—not blanket rejection.
Visibility and validation increase with institutional maturity.
The 10-Dan Framework does not assume AI is safe:
It assumes AI must be governed proportionally, transparently, and defensibly.
Progression is earned through institutional discipline—not technological enthusiasm.
✓ power governed, power deployed