Is your LiveChat chatbot EU AI Act ready?

LiveChat and Article 50, briefly

LiveChat and ChatBot are two products from the same company that often sit on one site together. LiveChat reaches a person, ChatBot answers on its own. From the outside you see the same window, so for Article 50 the first question is always: who gives the first answer, and does the visitor know?

What our scans found on LiveChat sites

We have not yet scanned enough sites where LiveChat 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 LiveChat

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.livechatinc.comapi.livechatinc.com

JS globals

window.LiveChatWidgetwindow.__lc

DOM selectors

#chat-widget-container

Where the AI notice goes in LiveChat

Settings, Chat, Greetings. Running LiveChat ChatBot? Put it in the bot welcome block, because that block wins over the greeting.

1. 1

   Open LiveChat and go to Settings, Chat, Greetings.

2. 2

   Edit the greeting that fires on your website.

3. 3

   Running ChatBot? Open the separate ChatBot dashboard.

4. 4

   Find the welcome block in your story and set the notice there.

5. 5

   Publish the story; a saved story is not yet live.

LiveChat documentation: www.livechat.com

What goes wrong most often with LiveChat

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.

• The notice is in LiveChat while ChatBot answers

  Teams running both products often only edit the LiveChat greeting. The moment ChatBot takes the conversation, the visitor sees the bot welcome block instead. Edit both.

• The bot only runs after hours, and so does the notice

  A common setup is people by day, bot by night. The notice then gets tied to the schedule and disappears during the day. That is error prone. Show it always; a visitor talking to a person is not harmed by one extra sentence.

• The bot name suggests a colleague

  ChatBot lets you pick a name and avatar. A bot called "Lisa" with no further explanation reads as a person. Our scanner flags that as a human persona, which is worse than no text at all.

Checking that it works in LiveChat

Test twice: once during office hours and once outside them, because on LiveChat your schedule decides whether ChatBot or a colleague sends the first message. Use an incognito window both times. Open the chat, read the name above the window and the first message, and check that something there says this is AI. If you edited the story in ChatBot, confirm that it was published as well; a saved story stays a draft and changes nothing for visitors.

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.

Frequently asked questions

Running another widget as well?

• Chatwoot and Article 50
• Freshchat and Article 50
• Landbot and Article 50