Skip to content
pitch
NL Nederlands
Menu

AI Governance and Policies

An AI policy your teams can follow, with an inventory of the tools you use, clear roles, an approval route for new tools and rules for acceptable use, linked to your GDPR work.

By
By , Attorney-at-Law
Published
Updated

The question

Your staff already use AI. Some tools were bought centrally, others were picked up by a team on a free account. Management wants to know which tools are in use, who decided, what data goes into them and who is accountable when something goes wrong. Customers and investors ask the same in their questionnaires.

The AI Act (Regulation (EU) 2024/1689) does not require a document called an AI policy. It does set duties that are hard to meet without one. A deployer of a high-risk system must use it in line with the instructions for use and assign human oversight to people who have the necessary competence, training and authority (Article 26(1) and (2)). Every provider and deployer must take measures to support the AI literacy of its staff (Article 4). And your role matters: a company that puts its own name on a high-risk system, or changes it substantially, can become its provider (Article 25(1)). See provider or deployer.

The GDPR runs alongside. A tool that takes decisions about people based solely on automated processing touches Article 22, and processing with new technologies that is likely to result in a high risk needs a data protection impact assessment under Article 35. A policy that ignores the GDPR side leaves half the questions open (where the two frameworks overlap).

Our approach

We start with what you have. In a short intake with IT, HR, legal and the business teams we list the AI tools in use or planned, what they are used for, which data they receive and who owns them. That inventory goes into our AI Act register, where each system gets its role and, where relevant, its risk class. If you need the full classification work, that is our AI Act compliance service; governance builds on it.

Then we write the policy around your organisation. A ten-person agency and an industrial group need different texts. We keep it short and practical:

  • Scope and roles. Who approves a new tool, who owns each system, who handles incidents, and who the staff can ask.
  • Approval of new tools. A short form and a decision route: what the tool does, what data it needs, the supplier's terms, and whether it changes your role under the AI Act.
  • Acceptable use. What staff may and may not put into a tool (client files, personal data, trade secrets), when output must be checked by a person, and how AI-generated content is labelled where the law asks for it.
  • Oversight and incidents. Who supervises systems where human oversight is required, how errors are reported and logged, and when a lawyer is called in.
  • Review. When the policy and the inventory are looked at again.

Where a system processes personal data, we link it to your records of processing and to a DPIA where one is needed, so one decision does not get made twice in two places. You and your team see and edit the registers with us in the client portal.

Our platform runs on our own AI, and a lawyer reviews everything it produces before it reaches you. That is also the standard we write into your policy: AI supports the work, and a named person decides.

For the legal background, read building an AI governance framework and our practical overview of the AI Act.

What you get

  • An AI policy in English or Dutch, approved by your management.
  • An acceptable use guideline for staff, short enough to read in ten minutes.
  • An approval form and decision route for new AI tools.
  • A register of AI systems with owner, role, purpose and data, kept with us.
  • A list of oversight roles and an incident procedure.
  • Links between AI systems and your GDPR records and assessments.
  • Where useful, a session to present the policy to your teams, which can be part of AI literacy training.

Limits apply. A policy describes how decisions are made; it cannot make a prohibited or unlawful use acceptable, and it does not replace the classification of each system. A policy also only works when people follow it, so we keep it short and assign each task to someone by name.

Next step

Send us a list of the AI tools your teams use, however rough, and any policy or guideline you already have. We will tell you what is missing and how much work it takes.

Schedule a meeting or contact us.

In short

  • An AI policy written for how your company works
  • Rules for acceptable use that staff can apply
  • An approval route for new AI tools
  • Named owners and human oversight per system
  • Linked to your records of processing and DPIAs
  • Reviewed by a lawyer

Our proprietary Legal Tech

Background reading

Let’s talk

Tell us what you’re working on. We’ll get the right person on it and reply within one business day. The first chat is free.

Schedule a meeting