Chat widget
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 Chatwoot.
Free · No account · Result in ~30s
Chatwoot is open source and often self-hosted, sometimes with your own language model behind it and sometimes with Captain, its built-in AI. Where it runs makes no difference to the obligation. What does matter: on a self-hosted install you can see exactly which message goes out first, and therefore exactly where the notice belongs.
We have not yet scanned enough sites where Chatwoot 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.
app.chatwoot.comwindow.chatwootSDKwindow.$chatwoot.woot-widget-holder.woot-widget-bubbleSettings, Inboxes, your website inbox, Welcome message. Enable the welcome heading too so it shows before opening.
Open Chatwoot and go to Settings, Inboxes.
Pick your website inbox and open the Configuration tab.
Fill in the welcome message with the AI notice.
Enable the welcome heading so the sentence shows before opening.
Running Captain or your own integration? Edit the first automated reply too.
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 Chatwoot: Chatwoot is open source and often self-hosted. The obligation does not depend on where it runs, only on whether an AI talks to people.
Chatwoot documentation: www.chatwoot.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.
A custom integration posts messages straight into the conversation and ignores the welcome message. We see tidy settings on top of a bot that introduces itself as a colleague. Put the sentence in the first message your code sends.
The heading is the only part visible before the visitor clicks. With it off, the notice arrives only after opening. That is still usable, but weaker than text already sitting next to the launcher.
If you keep a separate website inbox per brand or language, you have to walk all of them. Settings are per inbox, not per account. On Chatwoot this is the most common cause of a partial notice.
First count your website inboxes; with more than one you have to walk all of them, because settings live per inbox. Then open your site in an incognito window and check whether the welcome heading is visible before you click. Next start a conversation and watch which message arrives first: with a custom integration or Captain that comes through the API rather than from your settings. On a self-hosted install you can confirm this in your logs.
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.
Then it is definitely an AI interaction. Set the notice in the welcome message and in the first reply your integration sends, because the latter is what the visitor actually reads.
Same thing: put the notice in the first message Captain sends. The inbox setting covers the moment before that.
No. The AI Act looks at deployment towards people in the EU, not at your hosting choice. Self-hosting only makes it easier to see exactly which message leaves first.
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.