Zum Inhalt springen

Legal

AI transparency

We build AI systems for others. That is why we disclose where and how we use AI ourselves, and where we do not.

As of September 2026

On this website

No AI interacts with you on this website. The enquiry chat is a guided flow with fixed questions and answers we wrote ourselves. Graphics and animations are designed by us, not generated. The demonstrations on the project pages are simplified, pre-programmed examples.

Should we use AI-generated images or videos in future, we will label them visibly and permanently on the content itself.

In our work

We develop software with AI assistance. Concept, architecture decisions and review stay with people. No code goes into production without human review.

For client projects we use models from Anthropic and Azure OpenAI with processing in the EU. Client data is not used to train models.

In our products

Our systems follow the two-zone principle: the AI drafts, fixed rules decide what actually happens. Prices, commitments and contracts always rest with a person.

Every message written by an AI is labelled as such, as Article 50 of the EU AI Act requires.

No decisions about people

Our systems record details, make contact and book appointments. They do not make decisions about people, in particular in recruitment. Those remain with humans.

Competence

Everyone at Devonaut who works with AI systems is trained for it, as Article 4 of the EU AI Act provides. The owner is certified as an AI compliance officer (TÜV).

Questions

If you would like to know how AI is used in one of our systems, write to us at hallo@devonaut.ai.