How we make our content

AI helps us; a teacher decides. This page explains exactly where we use AI, who checks what, and who is responsible.

Updated on 11 October 2026

Who is responsible

Inburgering Oefenen is run by Samen Sterk in Taal (KvK 77533216). Editorial responsibility lies with Amber van der Zalm, a certified NT2 teacher. She decides what goes online.

About Amber

Where we use AI

Practice exercises and lessons: AI drafts a first version in DUO’s format. Amber checks every exercise and corrects it. Only then does it go online.
Guides and blog: AI helps with the first draft of the Dutch text. Every figure comes from an official source, with the date we consulted it. Amber reviews the text before it goes online.
English and Arabic translations: These are machine-translated from the reviewed Dutch text and have not been checked by a person. Every translated page says so at the top. If in doubt, read the Dutch version.
Grading of Writing and Speaking: A language model applies Amber’s marking scheme to your answer. For Speaking, ElevenLabs first turns your recording into text. Amber checks the gradings. A score is practice feedback, not an official DUO result.
Voices in the listening fragments: The voices are computer-generated (ElevenLabs). The speakers are fictional characters.
The coach in your dashboard: It works with fixed rules, not AI. It looks at your scores and tells you what to practise next.

Which data is used for grading is explained in our privacy policy. Privacy policy

What we don’t do

No made-up reviews. The reviews on our homepage are real Trustpilot reviews of our sister site knmoefenen.nl, quoted word for word.
No made-up authors. Anyone shown as a teacher on this site exists and can be looked up.
No figures without a source. If we can’t back a number up, such as a pass mark DUO does not publish, we don’t state it.

Spotted a mistake?

Let us know. Amber will look at it and we will fix it.

This page also describes how we meet the transparency rules of the EU AI Act (Article 50).