of AI
Control the use of AI.
Authority / evidence / oversight
AI and
Governance.
Two connected disciplines: governing AI so it is safe, lawful and accountable—and using AI carefully to make governance more informed, responsive and effective.
GOV / SYSTEM MAPHUMAN AUTHORITY / ACTIVE
of AI
Accountability↔Evidence
governance
Improve governance with AI.
People decideAI supportsEvidence endures
The essential distinction
Govern the technology.
Improve the institution.
Control how AI is chosen, built and used.
Define authority, acceptable use, risk controls, evidence and oversight throughout the AI lifecycle. The organisation remains accountable—even when technology or delivery is outsourced.
Explore governance of AI →Use AI to strengthen governance work.
Apply AI to find signals, review evidence and support oversight while preserving human judgement, due process, contestability and traceability.
Explore AI-enabled governance →One important boundary: AI-enabled governance is still an AI use case. It must pass through governance of AI before it can be trusted to support governance decisions.
Nine connected knowledge areas
From principle
to operating practice.
Each area answers a different governance question and produces evidence that the next decision can rely on.
01Govern AI
How do we keep AI lawful, safe, accountable and fit for purpose?
The management system around AI: decision rights, policies, controls and evidence spanning discovery, procurement, development, deployment, operation and retirement.
Open knowledge area →02Govern with AIWhere can AI strengthen oversight without becoming the decision-maker?
The responsible use of AI to support governance work: finding signals, reviewing evidence, monitoring obligations and helping accountable people make better-informed decisions.
Open knowledge area →03AccountabilityWho may decide what—and who remains answerable?
Align AI ambition with public value, organisational risk appetite and an operating model that gives leaders, clinicians, data owners and delivery teams explicit responsibilities.
Open knowledge area →04Risk and impactWhat could change, for whom, and how serious could the consequences be?
A proportionate method for classifying AI uses and assessing safety, rights, equity, operational, reputational and financial impacts before and after deployment.
Open knowledge area →05Clinical oversightWhat must people understand, review, override and escalate?
Design human involvement as a safety control with appropriate competence, time, information and authority—not as a disclaimer placed after an automated output.
Open knowledge area →06Data and privacyIs information used lawfully, appropriately and with enduring control?
Govern the data entering AI, the information it creates and the traces it leaves—from provenance and permitted use to retention, access, quality and records obligations.
Open knowledge area →07Ethics and lawWhat obligations and values must shape this use of AI?
Translate legislation, regulation, policy, professional duties and ethical commitments into concrete design requirements, operating controls and evidence.
Open knowledge area →08Third partiesWhat must suppliers prove before—and after—we depend on them?
Make AI-specific evidence, rights, responsibilities and exit conditions part of market engagement, evaluation, contracting and ongoing supplier management.
Open knowledge area →09Continuous assuranceHow do we know controls work and remain effective in production?
Create an evidence chain from requirements and pre-release evaluation to operational monitoring, independent review, incident response, improvement and retirement.
Open knowledge area →A lifecycle management system
Governance moves
with the use case.
Evidence accumulates from the first idea through every material change. The level of challenge increases with risk, autonomy and consequence.
Frame
Purpose, value, affected people
Classify
Impact, autonomy, sensitivity
Design
Controls, oversight, requirements
Assure
Evaluate, challenge, decide
Operate
Monitor, respond, improve
Retire
Exit, retain, learn
Healthcare lens
One governance system.
Multiple duties of care.
For health services, AI governance must connect—not compete—with clinical governance, patient safety, privacy, cyber security, information and records management, procurement, research ethics and quality improvement.
Build the operating system
Can every AI decision be defended?
We help turn governance obligations into practical authority, controls and evidence that move with delivery.
Start a conversation ↗