AI Governance
Control over how AI is used: ownership, policies, risk priorities and oversight.
The business question
People are already using AI, and the business needs control over what happens: what is in use, by whom, with what data, under whose responsibility, and on what rules.
This is the standalone governance service. Appropriate governance is also considered throughout every other engagement, and does not depend on buying it separately.
When this is useful
Recognisable situations, and the clients they suit.
Tools have been adopted team by team, with no register and no agreed rules.
You need to show who is responsible and what controls exist.
A pilot is about to become routine and the rules should come first.
Scope and exclusions
What the engagement covers, and what is separate work.
- An AI register: what is in use, by whom and for what
- An acceptable-use policy
- Assigned responsibilities
- Risk priorities
- A review process the business can run
- Legal opinions, including on data-protection law
- Certification or formal compliance attestation
- Specialist cybersecurity assurance or testing
- Operating or monitoring AI systems in production
What you receive
Concrete deliverables.
What is in use, by whom, with what data and under whose responsibility.
Rules that fit how your people actually work.
Named owners, the risks that matter most and a routine to keep it current.
How the engagement works
Stages, and who does what.
- 01Establish what is in useWe find out what tools and uses exist, including those nobody formally approved.
- 02Set the rulesWe draft the policy and responsibilities with the people who will operate them.
- 03Prioritise riskWe rank the risks by consequence and likelihood, so effort goes where it matters.
- 04Set up the reviewWe agree a review routine and hand over, so the controls stay current.
What we need from you
Access, information, decisions and participation. Where success depends on these, we say so at the start.
A candid account of what is used, including unofficial use.
People who will take the responsibilities and run the review.
Made by someone with the authority to make them.
Examples and limitations
Illustrative applications, not case studies. They show how a case is weighed and are not evidence of sector specialisation.
May help when the policy says what data may and may not be entered and people know why.
Limits a policy nobody follows is worse than none: it must match how work is done.
May help when a register lets you see duplication, cost and risk in one place.
Limits a register is only as good as the effort to keep it current.
May help when the need is a legal opinion, a certification or a security test.
Limits those need specialist providers. We can say when you need one.
Optional next steps
Where this leads, if you want it to.
Weigh the AI already in use against its value and its risks, and review it again at an agreed cadence.
The assessmentBuild the governance arrangements into the operating plan.
AI Strategy and AdoptionPut control in place.
Tell us what is in use and what worries you, and we will say what to set up first.