How we use AI: our AI-use policy

Why we publish this

We build learning tools, and we work with partners on AI literacy and applied AI in learning. Partners, learners and the organisations that fund projects should be able to check how we use AI ourselves.

Under Article 4 of the EU AI Act, as amended in 2026, organisations that build or use AI systems (providers and deployers) must take measures to support the AI literacy of their staff and of others who use AI systems on their behalf. The law does not require a specific level of AI literacy or a certificate. These rules are part of the measures we take. They are our own rules, not a certificate or a legal opinion. For a plain explanation of the law, see AI Act Article 4 explained.

This is not legal advice.

Where we use AI

These rules apply in three places:

In learning tools we build, when AI serves the learning goal. One example is spoken interaction with a virtual character in a VR role-play, where learners talk instead of clicking answer menus.
In research prototypes, such as an augmented-reality assistant with on-device person detection that we built as a feasibility study. It detects that a person is present; it does not identify who the person is.
In our own work, for example to help draft and edit texts, including articles on this website.

We integrate existing AI models into our work. We do not train our own foundation models.

We do not enter personal data of learners, clients, partners or project participants into an AI tool, unless a contract and a data processing agreement cover that use.
We never enter special-category data, such as health information or a person's migration status, into an AI tool.
We do not enter confidential information from clients or partners into public AI tools.
Learner data from our tools is not used to train a supplier's AI models without a written agreement.
When a feature needs data, we send only the data it needs, and we prefer EU-based or on-premises options where they exist.

How we handle personal data in general is described in our privacy policy.

Human review and oversight

A named person reviews every text, image or audio file made with AI help before we publish or deliver it. Crow Group B.V. takes editorial responsibility for what we publish.
In learning tools, AI supports learners and trainers. It does not decide on a learner's assessment: a person does.
We do not build AI that infers the emotions of people at work or in education. The AI Act prohibits this, except for medical or safety reasons.

This is what human oversight means for us: a person stays responsible for every decision that affects a learner, a client or a partner.

Labelling of AI-generated content

When people talk to an AI character in one of our tools, we tell them it is AI, unless that is obvious from the context.
If we use synthetic voices, images or video that could pass as real, we say so.
Articles on this website are published under human editorial responsibility, whether or not AI helped to draft them.

AI transparency, for us, means that nobody has to guess whether they are dealing with a person or with AI.

Choosing suppliers and models

Before we use an AI tool or model in a product, we check that its licence allows commercial use.
We check where a supplier processes data, and we sign a data processing agreement when personal data is involved.
We prefer models we can replace, and open-source models where they fit the task.
We keep a list of the AI tools we use. A new tool is added only after someone has checked it against these rules.

AI literacy in our own team

Everyone at Crow Group who uses AI tools at work takes part in a short session on using them well and safely, matched to their role. We keep an internal record of these sessions. The European Commission notes that no certificate is needed for this; an internal record is enough.

Version

Version 1, November 2026.

This policy is one of all our policies. Who we are: about Crow Group. Questions about this policy? Contact us.

Planning AI in a learning project? Ask about AI in your project. Or see AI literacy and applied AI.