Freshworks and the EU AI Act: does your chat widget tell people it's AI?
Freshworks spreads Freddy AI across Freshchat, Freshdesk, and the wider suite, so the 'chatbot' on a Freshworks-powered site may be one output of a platform-wide AI layer. For Article 50(1), the platform breadth doesn't change the question — it multiplies where you have to check it: every widget, portal, and messaging channel where Freddy answers an EU visitor first.
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 Freshworks
In a typical Freshworks setup, the surfaces that can carry (or fail to carry) the AI disclosure are:
- The Freshchat widget launcher and welcome message
- Freddy's bot persona (name, avatar, descriptor) in conversations
- Bot-flow answers and Freddy-generated responses before agent pickup
- Portal and in-product messaging entry points that share the same bot
Your Freshworks disclosure checklist
- List every customer-facing surface where Freddy responds (site widget, help portal, WhatsApp/social channels) and test each one's first message
- Set the bot persona to an explicit AI identity across all channels, not just the web widget
- Add the disclosure to the widget welcome and to the first bot-flow node so both paths carry it
- Verify channel-specific templates (e.g. WhatsApp) include the AI statement where the format allows
- Label agent-takeover events in the visitor-visible thread
- Export bot-flow definitions and screenshot each channel's opening state, dated, for your evidence file
Watch out for
Multi-product suites fail at the seams: the web widget gets fixed because it's visible to whoever ran the audit, while the help-portal bot and the WhatsApp channel keep running last year's configuration. Freshworks admins should treat the disclosure as a per-channel setting to sweep, not a single toggle — and re-sweep after suite updates, which can reset or reorganize bot settings.
Common questions
Freddy also drafts replies our agents send — is that a disclosure issue?
Agent-assist that a human reviews and sends is generally the weaker case; the visitor is genuinely corresponding with your agent. The hard requirement is the bot-led conversations. If Freddy answers autonomously anywhere, that's where the first-interaction disclosure must live.
We use Freshdesk's portal bot but not the site widget — still in scope?
Yes — the obligation follows the interaction, not the product name. A portal bot answering EU users conversationally needs the same first-interaction clarity as a homepage widget.
Check it in one scan.
DisclosureProof opens your site the way a regulator would, triggers your Freshworks widget, and records whether the AI disclosure actually appears — with timestamped evidence either way. Free scan at launch.
Join the launch list