Generative AI

Is your Voiceflow chatbot EU AI Act ready?

Article 50 asks that visitors know they are talking to an AI. Scan your site free, and read below exactly where that notice goes in Voiceflow.

Free · No account · Result in ~30s

Voiceflow and Article 50, briefly

Voiceflow gets used for two very different channels: a chat widget on your site and a spoken assistant on the phone. Same flow, two completely different notices. A text banner never reaches a caller, and a spoken line does nothing for a chat visitor. Run both channels and you have to handle it twice.

What our scans found on Voiceflow sites

We have not yet scanned enough sites where Voiceflow was the first widget we detected to say anything meaningful here. As soon as the sample is large enough it will appear, counts included. We would rather publish nothing than a percentage based on two sites.

How we detect Voiceflow

Our scanner looks for these signals on your public page. Two independent signals are required for a match, so a blog post that merely names the vendor does not count.

Script hosts
cdn.voiceflow.comgeneral-runtime.voiceflow.com
JS globals
window.voiceflow
DOM selectors
#voiceflow-chat

Where the AI notice goes in Voiceflow

A speak or text step as the very first step of your flow, plus the assistant description in the webchat embed config.

  1. 1

    Open your project in Voiceflow and go to the start of your main flow.

  2. 2

    Add a speak or text step as the very first step.

  3. 3

    Put the AI notice in it, matched to the channel.

  4. 4

    Open the webchat embed settings and fill in the assistant description.

  5. 5

    Publish and test both channels separately.

The sentence you paste

You are chatting with an AI assistant, not a person.

The welcome message only appears after opening. The banner snippet from the Compliance Kit covers the moment before that with a single script tag, and your AI transparency statement covers the question that comes after.

Watch out with Voiceflow: Using Voiceflow for a phone assistant? The notice has to be audible. A text banner never reaches a caller.

Voiceflow documentation: docs.voiceflow.com

What goes wrong most often with Voiceflow

Three mistakes we run into on sites that did set the notice. Each one costs points in the scan, and each one takes minutes to fix.

Checking that it works in Voiceflow

Test each channel separately, because one test says nothing about the other. For chat: open the embed in an incognito window and read the first message. For voice: call or start the assistant yourself and listen for the notice before you have said anything, and check that it stays audible if you start talking immediately. Also confirm the notice does not arrive only after intent recognition; you notice that when you have to ask a question before anything about AI is said.

After that you can scan again. Our scan looks at the same public page a visitor does, so what you see in an incognito window is roughly what we see.

Two dates

  • 2 August 2026: the transparency duty on interaction (Article 50) applies.
  • 2 December 2026: marking AI-generated content becomes relevant.

Frequently asked questions

Chat and voice in one project?

Set the notice in both channels: a spoken first line for voice, a first message for chat. Use a channel condition so you keep a single flow.

Does the typing indicator count as a notice?

No. The notice has to say in words that this is AI. An indicator, a robot icon or a blue outline are not text.

What if the user talks over the notice?

That is fine. The duty is to inform, not to obtain confirmation. You do not need an accept button; the sentence has to be present and understandable.

Running another widget as well?

Fix this for Voiceflow in about 15 minutes

The Compliance Kit contains a copy-paste disclosure banner (NL/EN/DE), an AI transparency statement, an AI system inventory and a step-by-step guide per widget.

See the Compliance Kit (€149)

Technical compliance tooling and templates. Not legal advice.