These pages answer the questions that come before the first scan: whether the rules reach you at all, what exactly has to be stated, from when, and how you evidence the state of your site.
Does this apply to me?
The entry point: company size, what kind of chatbot you run, being established outside the EU, and who carries the duty when the widget comes from someone else.
What exactly has to be stated?
The implementation: wording and placement of the disclosure in the chat, labeling of images, and where text is treated differently from images.
- How should an AI disclosure in a chatbot be worded?
- Is a notice in the privacy policy enough?
- Where in the chat window does the AI notice have to appear?
- How do I mark AI images in a machine-readable way?
- Is a visible watermark enough to meet the labeling requirement?
- Do I have to label AI-generated text on my website?
From when, and what else follows?
Deadlines, the Digital Omnibus, and the consequences question — answered through the mechanism rather than through numbers nobody can support.
Evidence and self-assessment
What stays visible from the outside once you have implemented: the findable transparency notice and a dated record of what was measured.
Why these fifteen
A question is on this list only if its answer either ends in a measurement we actually perform, or rests on a fact we can support from at least two independent sources. "What is the EU AI Act?" is not here: any language model answers that better than we would. What is worth writing down is the kind of question whose honest answer is: it depends on what is actually built into your site.
Check it on your own page
Run the free checkLegal note
This page provides general information on the EU AI Act — it is not legal advice and not a legal assessment of your individual situation (cf. German Federal Court of Justice, judgment of 9 September 2021, I ZR 113/20). It does not replace review by a lawyer. When in doubt, consult one.