Business guides – contracts, HR, AI & compliance

AI Governance – how companies build step-by-step governance for AI

Practical guide to AI Governance: inventory, classification, risk assessment, vendor due diligence, approval, AI literacy, and monitoring.
Published Mallbutiken
AI Governance – how companies build step-by-step governance for AI

By Mallbutiken · Facts verified October 1, 2026 · Approximately 9 minutes reading time

AI Governance is the organization's practical management system for AI: who is permitted to introduce an AI tool, how it is registered and classified, what risks are to be assessed, who approves the usage, and how it is followed up. An AI policy is an important part, but governance begins where the policy ends – in responsibilities, registers, controls, and decisions that can be demonstrated in retrospect.

Short answer: Start by taking an inventory of actual AI usage. Then, create an AI register, classify the use cases, conduct the appropriate risk and supplier assessments, document the approval, and establish recurring follow-ups. Avoid treating all AI systems as if they carry the same legal or technical risk.

What is AI Governance?

AI Governance is a collective term for an organization's rules, roles, processes, and documentation regarding AI. The purpose is to make usage controllable. It must be possible to answer questions such as: What AI systems do we use? Who owns them? What data is processed? What is our role under the AI Act? What risks have been accepted? When should the system be reviewed?

It is therefore useful to distinguish between three levels. Policy describes the overall rules of the game. Process describes how a new use case goes from idea to approved operation. Evidence is the documentation showing which assessments and decisions have actually been made.

Read the guide on AI policies for companies first if you lack internal ground rules.

A practical AI Governance model in eight steps

  1. Inventory. Map out tools, integrations, built-in AI features, APIs, and local models already in use.
  2. Register. Assign each use case an ID, system owner, business owner, supplier, purpose, and status.
  3. Classify. Assess the organization's role and whether the usage may be subject to prohibitions, high-risk rules, transparency requirements, or other regulations.
  4. Assess data and risk. Document personal data, trade secrets, security, error risk, discrimination, human oversight, and operational impact.
  5. Review the supplier. Check terms, training on customer data, storage, subcontractors, security, incidents, and exit strategies.
  6. Decide. Define who is authorized to approve, set conditions for, pause, or reject a use case.
  7. Implement and train. Link access and training to actual roles and usage.
  8. Follow up. Set dates for reassessment and triggers for earlier review, such as major version updates, new data sources, or changed areas of use.

What should an AI register contain?

Field Example of question
Owner Who is responsible for the system and the business value?
Purpose What will the AI actually be used for?
Data Is personal data, sensitive information, or trade secrets processed?
AI Act role Is the organization a provider, a deployer, or another actor in this situation?
Risk class What legal and internal risk classification has been made?
Human oversight Who reviews results before an important decision or external use?
Supplier What terms, subcontractors, and security controls apply?
Status Idea, pilot, approved, conditional, paused, or decommissioned?
Review When should the usage be reassessed?

A register is particularly valuable because AI features often appear inside systems that the organization is already using. The inventory should therefore not just ask "Do you use ChatGPT?" but also capture, for example, AI in CRM, HR systems, analytics tools, customer service, and coding platforms. See the separate guide to AI registers for fields and workflows.

Classification and risk assessment are two different things

A common mistake is to create a proprietary color scale for risk and treat it as the legal classification according to the AI Act. The internal model can help you prioritize controls, but it does not replace the assessment of the regulation's categories and obligations. Also read AI Act risk classification – when is an AI system high-risk?.

A practical risk assessment can cover, among other things, data protection, information security, hallucinations and data quality, discrimination, impact on humans, transparency, copyright, supplier dependency, and continuity. When personal data entails a likely high risk, a DPIA under GDPR Article 35 may also need to be considered.

Review the AI supplier before usage becomes critical

The supplier review should answer more than just whether the service "is GDPR compliant." Check where data is processed, whether customer data is used for training, how data is deleted, which subcontractors are used, what security functions are in place, and what happens in the event of an incident or supplier change.

For a practical procurement and approval process, use the checklist for AI supplier assessment. If the supplier processes personal data on your behalf, a data processing agreement (DPA) may be needed. See the guide on DPA. If the AI feature is part of a cloud service, the SaaS agreement is also part of the governance.

Who should decide on AI?

Avoid the model where "IT is responsible for AI" without clearer accountability. A use case can involve IT, data protection, HR, legal, information security, and business owners simultaneously. Therefore, appoint at least one business owner and one system owner, and define which functions must be involved in the event of increased risk.

A simple approval model can have three outcomes: approved, approved with conditions, or not approved. Document the conditions, such as that sensitive personal data may not be entered, that output must be reviewed, or that usage may only occur in a specific enterprise version.

AI literacy must be linked to the role

Article 4 of the AI Act requires that providers and deployers take measures to support the development of AI literacy for staff and others who use AI systems on their behalf. Consideration should be given to, among other things, technical knowledge, experience, training, usage context, and the individuals affected.

This supports role-based training: basic rules for broad usage, in-depth knowledge for system owners, and specialized training for, for example, HR, developers, or individuals approving new systems.

Checklist for a functional AI Governance system

  • Is there a responsible owner for AI Governance?
  • Do you have an updated AI system register?
  • Is there a process for new tools and new use cases?
  • Do you distinguish internal risk class from legal classification?
  • Are GDPR/DPIA and other relevant requirements screened?
  • Are external AI suppliers reviewed before approval?
  • Are approvals and conditions documented?
  • Is human oversight clearly defined?
  • Is training adapted to the role and risk?
  • Are there dates and triggers for reassessment?
AI Governance package with AI policy, register, risk assessment, and governance

Build a cohesive AI Governance workflow

Mallbutiken's AI Governance Package 2026 includes an AI policy, AI system register, risk assessment, supplier assessment, impact assessment, and governance/approval in Word, PDF, and Excel. Price in store: 199 SEK.

View the AI Governance package

Frequently asked questions

Is AI Governance the same thing as an AI policy?

No. The policy sets the rules of the game. Governance also covers inventory, classification, roles, risk assessments, supplier control, decisions, and follow-up.

Must every AI tool undergo the same review?

No. Controls should be proportionate to usage and risk. A simple low-risk tool does not normally need the same depth of review as a system that affects people or critical processes.

Do small companies need AI Governance?

If AI is used in important workflows, basic governance is valuable even in small companies. The model can be simpler, but it should still be clear which tools are approved and who makes the decisions.

Practical templates: AI agent controls – delegated authority, human approval and rollback. Assess the scope against your AI use, risks and applicable requirements.

Templates in this article

Explore the topic