Governance that works inside the organisation, not on a shelf
AI extends what data can do. It does not replace the foundation underneath it. Governance here is proportionate to what an organisation actually does: strong where the consequences are significant, simple where the risks are low.
What this covers
The work, in practice
AI governance diagnostics
An assessment of where AI is already in use, often before anyone has formally decided to adopt it: what it is doing, what data feeds it, which vendors are involved, what decisions it influences, where human oversight sits, and what governance already exists versus what is missing. A realistic starting point rather than a theoretical one.
Building practical AI governance
Frameworks and policies, use-case registers, risk identification, clear roles and decision rights, approval processes, human oversight, vendor governance, and the evidence trail behind all of it - proportionate to what the organisation actually does.
How clients start
Say what you need, in your own words
You should be able to come to Griffinity and simply describe the problem. For this area, that usually sounds like one of these.
- Help us establish practical AI governance, and build the frameworks with us rather than for us
- Help us understand our vendors, and whether AI is touching our data anywhere we haven't accounted for
- Review what we've already implemented
What you leave with
Advice becomes deliverables
The exact output depends on the scope. The aim is never simply to produce documents. It is to build foundations that reflect how the organisation actually collects and uses data.
- AI use-case inventory and governance gap assessment
- AI governance framework and policies
- Risk identification and proportionality model
- Roles, decision rights and approval processes
- Human oversight and escalation points
- The evidence trail that demonstrates responsibility
How engagements work
From diagnosis to ongoing practice
Griffinity supports the full journey, and you can engage at any single point in it or move through the whole arc.
Diagnose
Understand where you are.
Audit
Identify the gaps and the risks.
Advise
Determine what is actually needed.
Build
Create the foundations and the documents.
Implement
Put practical process into operation.
Govern
Establish accountability and oversight.
Review
Assess and improve what is already live.
Teach
Build knowledge in the people who live with it.
Evolve
Adapt as the organisation and its AI change.
Governance is not a once-off project. We walk through implementation, review what has gone live, assess new vendors and new use cases, and help capability mature as your technology does.
Also part of the same foundation
Start with the decisions your organisation needs to make
Tell us what is changing, what feels unclear, and what leadership needs to be able to approve or explain.