Userlike and the EU AI Act: does your chat widget tell people it's AI?
Userlike, a Cologne-based live-chat platform built for the EU market, lets you create a chatbot the same way you create a human operator: Team > Chatbots > Add chatbot asks for a first name, last name, and operator group — the identical fields used for a real teammate. That symmetry is convenient for admins and is precisely the Article 50(1) risk sitting in one setting screen: nothing in the setup flow forces the bot's identity to read as non-human.
The rule itself is short. Article 50(1) of the EU AI Act requires AI systems that interact directly with people to be designed so users know they're dealing with AI — no later than the first interaction, in a clear and distinguishable way. The exception for cases where it's "obvious" is narrow: a natural-language customer-service bot doesn't qualify just because it has a robot icon. A line in your terms of service doesn't satisfy it either. And it's easy to enforce, because a regulator can simply open your site and start a chat.
Where the disclosure lives in Userlike
In a typical Userlike setup, the surfaces that can carry (or fail to carry) the AI disclosure are:
- The chat widget launcher and the operator identity shown as a conversation starts
- The chatbot's configured first name, last name, and profile picture — set through the same fields as a human operator
- The chatbot's first automated message once a visitor starts chatting
- The point where a chatbot conversation is picked up by a real operator from the same operator group
Your Userlike disclosure checklist
- Open your site in a private window and start a chat: does the operator identity or the first message make clear you're talking to AI?
- Review your chatbot's first/last name and profile picture in Team > Chatbots — because these are the same fields as a human operator's, nothing stops an accidental human-presenting setup
- Add an explicit AI statement to the chatbot's first message rather than relying on the name alone
- If the chatbot shares an operator group with human agents, check that visitors can tell which kind of "operator" picked up their chat
- Verify the disclosure holds on Userlike's mobile SDK/app integrations, not just the website widget
- Screenshot the first-interaction state and export the chatbot's profile settings with a date
Watch out for
Userlike's chatbot setup is deliberately built to feel like adding a teammate — same fields, same operator group, same profile picture upload. That's good onboarding design and a silent compliance gap: an admin who names the bot "Lisa" with a stock headshot has, without any dishonest intent, built exactly the kind of non-obvious AI interaction Article 50(1) targets.
Common questions
We use the chatbot only to route to the right department, not to answer questions — is that in scope?
Pure routing with no generated answers is a weaker case for AI-interaction disclosure, but if the bot uses NLP/AI to interpret free-text requests rather than fixed buttons, treat it as in scope. When in doubt, a one-line disclosure costs little.
Is a "bot" status icon next to the name enough, since Userlike shows operator status?
Treat status icons as a nice-to-have, not the disclosure itself — the safer pattern is an explicit line in the bot's first message, since icon conventions aren't something every visitor reliably reads.
Check it in one scan.
DisclosureProof opens your site the way a regulator would, triggers your Userlike widget, and records whether the AI disclosure actually appears — with timestamped evidence either way.
Scan your site free