Guide AI

AI Governance for UK Organisations: A Practical Framework

Many organisations already use more AI than their IT team knows about. Staff are pasting customer data into ChatGPT, vendors have quietly enabled AI features, and no one has written a policy. This guide covers how to map what you have, classify the risks, and put the controls in place before something goes wrong.

24 August 2026
Updated 18 November 2026
9 min read
Key takeaways
  • Start with an inventory. You cannot govern what you have not catalogued. Shadow AI is widespread, and vendor AI is often enabled by default in tools you already pay for.
  • Classify tools by the data they process and the decisions they influence. A low-risk AI tool and a high-risk one need different controls.
  • An AI acceptable use policy does one job: tell staff what they can use, what data they cannot enter, and what review is required before acting on AI outputs.
  • Under UK GDPR, as amended by the Data (Use and Access) Act 2025 (Articles 22A to 22D, in force since 5 February 2026), solely automated decisions with legal or similarly significant effects on individuals need safeguards, and those based on special category data are more tightly restricted. HR screening tools, credit decisions and insurance pricing can fall into this. Having a person somewhere in the process does not on its own make a decision less than solely automated: the human involvement must be meaningful, by someone with the authority and information to change the outcome, and documented.
  • Human oversight means a human who genuinely evaluates the AI output before a consequential decision is made. Rubber-stamping is not oversight.
  • The EU AI Act can reach UK organisations, but having EU customers or employees does not on its own trigger every obligation. Article 2 looks at concrete tests: placing an AI system on the EU market, using it from an establishment in the EU, or the system's output being used in the EU. High-risk AI obligations apply from December 2027.
  • ISO 42001 certification is not necessary for most UK SMEs unless a customer or contract requires it. Even without certification, it can be useful to adopt measures from the standard.

What AI governance actually is

AI governance is the combination of policies, controls, and accountability structures that determine how your organisation uses AI tools, which data those tools can access, how AI-influenced decisions are made and reviewed, and who is responsible when something goes wrong.

It is not an ethics statement, an AI strategy document, or a list of principles. Those are starting points for conversations, not controls. Governance means specific rules that apply to specific tools, specific accountability for specific decisions, and a review process that actually runs.

Many UK organisations at the start of 2026 have none of this. They have staff using ChatGPT, Copilot, Gemini, and a dozen SaaS tools with AI features enabled, with no policy, no training, and no visibility into what data is being processed. The governance gap is not theoretical: it is a UK GDPR exposure, a reputational risk, and, for organisations with EU operations, an EU AI Act compliance question.

Establish ownership first

Before inventorying tools or writing policy, decide who is accountable for AI governance. Without a named owner and clear decision rights, the inventory does not get maintained, the policy does not get enforced, and incidents do not get escalated. In most UK organisations the structure looks like this:

Role AI governance responsibility
Board / Executive Set risk appetite for AI use. Receive quarterly reporting on AI risk posture. Approve high-risk AI use cases.
AI Governance Owner (IT lead, DPO, or vCISO) Maintain the AI tool register. Approve or reject new tool requests. Coordinate DPIA completion. Own the acceptable use policy. Report incidents upward. Run the review cadence.
DPO / Privacy Lead DPIA assessments. Advise on the automated decision-making rules (UK GDPR Articles 22A to 22D). Review vendor DPAs. Handle subject access requests involving AI.
IT / Security Assess AI tool security controls. Manage access permissions for enterprise AI tools. Detect shadow AI. Monitor for data exfiltration. Handle AI-related security incidents.
Legal / Compliance Review AI-related vendor contracts. Advise on sector-specific regulatory obligations. Track EU AI Act developments for relevant use cases.
Business / Department Owners Own AI tool use cases within their function. Ensure staff follow the acceptable use policy. Escalate new AI tool requests through the formal approval process. Own human oversight within their workflows.
End Users Follow the acceptable use policy. Use only approved tools. Report incidents and shadow AI use. Complete AI awareness training.

In a smaller organisation, several of these roles will sit with the same person. That is fine. What matters is that the responsibilities are explicitly assigned, not that each has a dedicated headcount. A 50-person business with a named AI governance owner, a defined escalation path, and a quarterly review meeting has more functional governance than a large organisation with a governance framework document and no one running it.

The ten steps

Once someone owns AI governance, the work comes down to ten steps. We cover them in three companion guides.

AI inventory, risk classification and acceptable use policy covers steps 1 to 3: finding every AI tool in use, including the ones staff signed up for themselves, classifying each tool by the data it handles and the decisions it influences, and writing a policy staff can read in five minutes.

AI data protection, human oversight and security controls covers steps 4 to 6: your UK GDPR obligations, including the automated decision-making rules in Articles 22A to 22D and DPIAs, human review that changes the outcome, and controls for risks such as prompt injection and data leakage.

AI vendors, incidents, the EU AI Act and ISO 42001 covers steps 7 to 10: due diligence before you approve an AI tool, what counts as an AI incident and how to respond, when the EU AI Act reaches a UK organisation, and whether ISO 42001 is worth adopting.

Your first 90 days

Many organisations find that AI governance feels overwhelming in the abstract but manageable when broken into a sequenced 90-day programme. The goal at the end of 90 days is to be able to answer: what AI do we use, who owns it, what data does it process, what risks exist, and what controls are in place?

D1
Days 1 to 30: Discover
Appoint your AI governance owner. Survey department heads to identify AI tools in use. Scan SaaS estate for shadow AI. Conduct staff survey to surface personal account use. Build the first version of the AI tool register. Identify any use cases that are obviously high risk and need immediate attention.
D2
Days 31 to 60: Control
Classify each tool in the register by risk level. Approve or put on hold any unapproved tools. Draft and publish the AI acceptable use policy. Complete vendor DPA checks for all medium and high-risk tools. Begin DPIAs for high-risk processing. Establish the human oversight process for any AI-influenced consequential decisions. Define and communicate the AI tool approval process.
D3
Days 61 to 90: Operationalise
Run AI awareness training for all staff. Finalise the AI tool register with all controls documented. Establish the quarterly review cadence. Report AI governance status to board or senior leadership. Test the AI incident management process with a tabletop exercise. Document evidence of controls for audit readiness.

Governance cadence

AI governance is not a project with an end date. The tools change, vendor terms update, new capabilities emerge, and your organisation's use evolves. Build a review cadence that keeps the programme current without consuming disproportionate time.

📋
Every quarter: AI tool inventory review
Check for new tools being used or requested, vendor changes to terms or AI features, and any incidents involving AI outputs. Update the tool register. AI capabilities and vendor terms move faster than annual reviews can track.
📄
Every six months: policy and training review
Update the acceptable use policy to reflect approved tool changes and any new guidance from the ICO or sector regulators. Refresh awareness training. If your organisation has grown or changed significantly, reassess whether any new AI use cases have emerged that are not covered.
📊
Annually: DPIA review and board reporting
Review DPIAs for high-risk AI tools, particularly where the tool's functionality or your use of it has changed. Report AI governance status to the board or senior leadership as a standing item, covering tool inventory, any incidents, regulatory changes, and the EU AI Act timeline if relevant to your organisation.

Assign a named owner for AI governance, whether that is your DPO, IT lead, or a vCISO carrying the function. Without a named owner, the review cadence does not happen.

Minimum viable AI governance for SMEs

If your organisation does not have a dedicated compliance team, start here. These eight actions give you meaningful governance without requiring a large programme.

📋
1. Build an AI register: know what you are using
List every AI tool in use across the organisation. Include tools staff use personally for business tasks. This is your baseline.
📄
2. Write an acceptable use policy: tell staff what they can and cannot do
Approved tools, prohibited data inputs, output review requirement, no personal accounts. One page, plain language.
⚖️
3. Classify your highest-risk use cases: find the ones that matter most
You do not need to risk-classify everything on day one. Identify the two or three AI use cases with the most data sensitivity, the most impact on individuals, or the most regulatory exposure, and address those first.
🔍
4. Check your vendors: know where your data goes
For every approved AI tool processing personal data, confirm a DPA is in place, check where data is processed, and establish that the vendor does not train on your data by default.
👤
5. Build in human review for consequential decisions
We recommend that any AI-influenced decision with a significant effect on an individual goes past a person who genuinely evaluates the output and can change it before the decision is final. This is good governance practice rather than a general legal requirement: UK GDPR Articles 22A to 22D set specific rules and safeguards for decisions based solely on automated processing. Name the reviewer and document the process.
📋
6. Run a DPIA where UK GDPR requires it
If you use AI to process special category data, profile individuals at scale, or make automated decisions with significant effects, that processing is likely to trigger the UK GDPR Article 35 DPIA requirement. Assess it against the ICO's guidance for each use case and run the DPIA before deploying the tool, not after.
🔒
7. Apply basic security controls: check permissions, log usage
Review access permissions for any AI tool connected to your data. Ensure logs capture AI tool usage. Establish how staff report an AI-related incident.
🗓️
8. Name an owner and set a review date
Someone is accountable. The AI register and policy have a next-review date. Without these two things, everything else fades within six months.
Ready to implement a specific use case?

This guide covers ongoing governance: the register, the policy, the controls that apply across every AI tool in use. When you are ready to deliberately build AI into a specific process, from business case through pilot, testing, and a measured result, our AI Implementation Guide covers that project end to end. It assumes the governance foundation in this guide is already in place and builds on it.

Common questions

Does the EU AI Act apply to UK organisations?

Under Article 2, the EU AI Act reaches, among others, providers placing AI systems on the market or putting them into service in the EU, wherever they are based; deployers established in the EU; and providers and deployers outside the EU where the output produced by the AI system is used in the EU. Having EU customers, operations or employees does not on its own trigger every obligation: you need to look at each AI system and your organisation's role in relation to it. Under the revised implementation timeline, Annex III high-risk AI rules apply from 2 December 2027; certain high-risk systems in regulated products apply from 2 August 2028. Even UK organisations not directly in scope may face contractual requirements from EU customers or partners who are in scope and need their supply chain to demonstrate compliance.

Sources

AI Governance Assessment

Not sure where your AI governance stands?

We assess your AI inventory, risk classification, privacy obligations, security controls, and governance structure. Deliverable: a gap assessment and prioritised 90-day roadmap.

  • ✓ What AI tools are in use and what data they process
  • ✓ Which use cases require additional controls or a DPIA
  • ✓ Where security controls for AI need to be strengthened
  • ✓ Whether EU AI Act or sector regulation applies
  • ✓ Who owns governance, approves tools, and handles incidents
  • ✓ Prioritised 90-day roadmap to close the gaps

Disclaimer: This article is for general informational purposes only and does not constitute legal, regulatory, or professional advice. Cyvra makes no warranty as to the accuracy or completeness of this content, which may not reflect the most current regulatory developments. Readers should seek independent legal and regulatory advice appropriate to their specific circumstances. Cyvra accepts no liability for any loss arising from reliance on this content.