Glossary
AI disclosure (AI Act, Article 50) – what does it mean?
Definition
AI disclosure (AI Act, Article 50): AI disclosure is the mandatory notice that the customer is talking to artificial intelligence and not to a human.
Article 50 of the EU AI Act has required it since 2 August 2026.
How it works
The rule is simple: the visitor must see that they are talking to a machine. The disclosure appears in two places: in the header of the chat window and in the first message. The first message introduces the assistant and indicates that the answer may be wrong.
Article 50 of the EU AI Act requires this. The disclosure is built in and cannot be switched off, so business leaders do not need to set it up separately.
An example at a company
On a service provider’s website, the widget header reads: “Artificial intelligence, not a human”. The first message introduces the assistant and indicates that it may make mistakes. When the visitor asks whether they are talking to a human, the assistant confirms that it is artificial intelligence and, on request, directs the question to a colleague.
What this means for business leaders
Ask the provider where the disclosure is shown and whether it can be switched off. A disclosure that can be switched off is a bad sign: the obligation may be breached.
If the disclosure is missing or hidden, the transparency obligation is not met, and your company is responsible for it. The right answer: the disclosure is always visible and cannot be switched off.
How we do it
Our disclosure appears in the widget header and in the first message, in line with Article 50 of the EU AI Act; it cannot be switched off. The data stays in the EU, including the model calls, and the retention period is 90 days. Details on our data security page.
What should you do now?
Do you have a question about your own website or online store? Request a demo or see our pricing.