Generative AI
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 Botpress.
Free · No account · Result in ~30s
Botpress is a developer platform, not a ready-made widget. You build the flow yourself, attach a language model and embed the webchat with a config object in your own page. That gives you full control over the first message, and it also makes you the one accountable for that first line being there.
We have not yet scanned enough sites where Botpress 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.
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.
cdn.botpress.cloudfiles.bpcontent.cloudmessaging.botpress.cloudwindow.botpressWebChatwindow.botpress#bp-web-widget-container.bpw-widget-btn#bp-widgetThe first node of your main flow, plus the webchat config, so it shows before the first message arrives.
Open your bot in Botpress Studio.
Go to the main flow and select the start node.
Add a text message card with the AI notice as the first card.
Open the webchat config and set the same sentence in the description or opening message.
Republish the bot; the embed loads the published version.
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 Botpress: Botpress bots are almost always generative. Beyond the interaction notice, decide now how you will mark AI-generated output (2 December 2026).
Botpress documentation: botpress.com
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.
Instructing the model to introduce itself as an AI is no guarantee that it will, and the visitor never reads that prompt anyway. Put the sentence in the flow as a fixed text message, not as an instruction to the model.
The webchat embed points at a specific bot version. We regularly see an updated flow in Studio while the site still loads the previous build. Always verify in an incognito window after publishing.
Botpress bots generate text. From 2 December 2026 marking AI-generated content becomes relevant. Handle only the interaction notice now and you will redo this work in a year. Write the approach into your AI transparency statement.
First check in your HTML which bot id and which version the webchat embed loads, and compare that with the version you just published. Then open your site in an incognito window and start a conversation: the fixed text message from your first node should appear before the model generates anything. Next, ask whether you are talking to a person and read the answer; with a generative bot that is the only way to see what your system prompt actually does with it.
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.
Makes no difference to the obligation. Self-hosted, you set the notice in the same first node; only where you publish differs.
Building the bot for yourself makes you a deployer. Building it for clients and placing it on the market can make you a provider, which carries more duties. Put that to a lawyer.
Technically yes, but a variable that stays empty produces an empty first bubble. We see that more often than you would expect. Hardcode the sentence, or test the empty state explicitly.
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.