EU AI Act · Article 50(1) · Chat widgets

Zendesk and the EU AI Act: does your chat widget tell people it's AI?

Applies from 2 Aug 2026Fines up to €15M / 3% turnoverLast reviewed July 2026

Zendesk's AI agents (the evolution of Answer Bot) typically sit in front of your ticket queue inside the Web Widget or messaging experience, deflecting questions before a human ever sees them. Because the same widget carries both bot and human conversations, the Article 50(1) question on a Zendesk site is rarely "is there AI?" — it's "can a visitor tell which one they've got, from the first message?"

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.

Who's on the hook? Responsibility is split: Zendesk carries provider-side duties for the system itself, but how the widget is configured and presented on your site is your deployment. If the disclosure setting exists and isn't enabled — or the notice is hidden — that gap is yours.

Where the disclosure lives in Zendesk

In a typical Zendesk setup, the surfaces that can carry (or fail to carry) the AI disclosure are:

The Web Widget / messaging launcher and its greeting state
The AI agent's persona: display name, avatar, and any bot badge shown in the conversation
The first automated reply — often an answer or article suggestion generated before any human touches the thread
Escalation and transfer messages when the conversation moves from the AI agent to a support agent

Your Zendesk disclosure checklist

Trigger the widget on your live site and note the first automated message: does it identify itself as AI or an automated agent?
Review the AI agent's persona settings — avoid human names and photos without an explicit AI descriptor
Add a plain-language AI statement to the first response, not only to a help-center page
Test the flow that starts from an email-to-messaging or help-center entry point, since greetings can differ per channel
Confirm the transfer-to-human message announces the change of participant
Export or screenshot the bot persona configuration and greeting flows with a timestamp for your records

Keep proof. On a complaint, an authority will want to see that the disclosure was there and how it was designed. Keep timestamped screenshots of the first-interaction state, your widget configuration, and a record of when each was last changed — that evidence file is the difference between "we comply" and "we can show we complied."

Watch out for

Zendesk deployments are channel-sprawl machines: the same AI agent can answer in the web widget, in-app messaging, WhatsApp, and email-adjacent flows, each with its own greeting configuration. Teams commonly fix the disclosure in the web widget and forget the other channels — and Article 50 doesn't care which channel the EU user happened to use.

Note: vendor interfaces and setting names change. This page describes where disclosure surfaces typically live in Zendesk as of July 2026 — verify the exact toggles in your own Zendesk workspace and against Zendesk's current documentation.

Common questions

Our Zendesk bot mostly suggests help-center articles — is that in scope?

If it interacts conversationally with users, treat it as in scope and disclose. The cost of a one-line statement is trivial next to arguing about whether article suggestion is 'interaction' after a complaint.

Does Zendesk handle this for us?

Zendesk provides the configuration surfaces; what visitors actually see on your site is your deployment. Split responsibility means you can't outsource the visible notice.

Check it in one scan.

DisclosureProof opens your site the way a regulator would, triggers your Zendesk widget, and records whether the AI disclosure actually appears — with timestamped evidence either way. Free scan at launch.

Join the launch list