Do I need to turn off my clinic's chatbot?

No. The obligation is about transparency, not the technology — the chatbot just needs to disclose, in its first message, that it's an automated system.

What if the notice only appears after several messages?

That doesn't meet the obligation. The law requires the notice from the start of the interaction, not partway through.

Does Your Clinic's Booking Chatbot Comply with AI Act Article 50?

If your clinic has a chatbot on its website — for booking appointments, answering common questions, or initial triage — there's a new legal obligation in force since August 2, 2026, with no transition period: patients need to know, from the very first interaction, that they're talking to an artificial intelligence system.

This is Article 50 of the AI Act, the EU's regulation on artificial intelligence. And unlike many regulatory obligations that only apply to large companies, this one has no size exemption — it applies to a two-person clinic just as much as to a hospital group.

What the law requires, specifically

The rule is simple to state, but easy to get wrong in practice: any AI system designed to interact directly with people must disclose that fact from the start of the first interaction, in a way that is clear and distinguishable — not buried in terms and conditions, not tucked away in a footer.

It doesn't need to be an alarming notice. It can be as simple as the chatbot itself saying, in its first message: "I'm a virtual assistant for [Clinic]. I can help you book your appointment." What the law doesn't allow is for the patient to be left believing they're talking to a human receptionist when they're not.

Where most clinics fail without knowing it

In practice, we've seen three situations that go unnoticed:

The notice exists, but only after the patient has already sent several messages. The law requires disclosure from the start of the interaction, not partway through the conversation.
The chat widget is powered by an AI product (not an on-call receptionist), but nothing on the site distinguishes it from a normal human chat. If the patient never realizes they're talking to an automated system, the obligation isn't met — even if the widget "looks like" just another WhatsApp button.
The notice is hidden inside the chat window itself, but only shows up if the patient clicks "About" or similar. The law requires it to be understood without effort, not just present somewhere.

What to do now

You don't need to hire a lawyer to fix this — in most cases it's a few minutes' change to the chatbot's welcome message. The practical steps:

Confirm your clinic's chatbot is actually an AI system (not a slow-responding human chat disguised as a bot) — if it's human, this obligation doesn't apply.
Check the first message a patient sees when they open the chat. Does it clearly say it's an automated assistant?
If not, add that line to the widget's welcome message — usually configurable without touching any code.
Save a dated screenshot as a record that the notice was already in place from now on.

A free check, in under a minute

If you're not sure whether your clinic's website shows this notice correctly — or if you'd like a broader diagnostic of what needs attention on the site (not just this, but also what's holding you back in Google search) — DENVORA SEO's free diagnostic scans the site in seconds and shows the specific points to fix, in plain language, no technical jargon.

Note: this article is general information, not legal advice. If you have specific questions about your clinic's compliance, speak with a lawyer specializing in data protection and AI regulation.