Gorgias and the EU AI Act: does your chat widget tell people it's AI?
Gorgias's AI Agent is built for Shopify-style ecommerce support and ships with a default identity of "AI Agent Bot" — but the platform explicitly invites you to rename it, pick a custom avatar, and choose a personality preset (Friendly, Professional, Sophisticated, or Custom). That's a lot of humanizing knobs on a system that, by default, answers customers before any human touches the ticket — precisely the Article 50(1) fork: keep the honest default name, or rebrand it and now owe an explicit disclosure somewhere else.
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 Gorgias
In a typical Gorgias setup, the surfaces that can carry (or fail to carry) the AI disclosure are:
- The storefront chat widget launcher and its opening message
- The AI Agent's configured identity — display name and avatar, set per brand in Gorgias
- The personality/tone preset applied to its replies (Friendly, Professional, Sophisticated, Custom)
- The escalation point where a ticket moves from the AI Agent to a human support agent
Your Gorgias disclosure checklist
- Open your storefront in a private window and message support: does the first reply identify itself as AI before or as it answers?
- Check the AI Agent's configured name and avatar — the default "AI Agent Bot" self-discloses; a rebrand to a human-style name/photo needs an explicit AI line to compensate
- If you set a Custom personality/tone, re-read the instructions for anything that nudges the agent toward sounding human rather than merely friendly
- Add the AI statement to the opening message template so it survives future rebrands
- Confirm the handoff-to-human moment is visible to the customer as a participant change, not just an internal ticket reassignment
- Screenshot the widget's first-interaction state and export the AI Agent's name/avatar/personality settings with a date
Watch out for
Gorgias gives ecommerce brands real design freedom over the AI Agent's identity — good brand practice and a compliance trap in the same feature. The moment "AI Agent Bot" becomes "Sam from Support" with a human-style avatar, the disclosure that used to be free (the default name) has to be rebuilt explicitly somewhere a visitor actually reads it.
Common questions
Our default is "AI Agent Bot" — are we already covered?
That default name does real disclosure work, but don't rely on wording alone — pair it with a plain first-message line, since a small header label is a weaker position than an explicit statement if it's ever examined.
We use the Custom personality with very casual, human-sounding instructions — is that a problem?
Tone isn't the issue; identity is. A casually-toned bot that's still clearly labeled AI is fine. The risk is specifically pairing a human-sounding name/avatar with no AI cue anywhere in the conversation.
Check it in one scan.
DisclosureProof opens your site the way a regulator would, triggers your Gorgias widget, and records whether the AI disclosure actually appears — with timestamped evidence either way.
Scan your site free