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

LivePerson's Conversational Cloud is a build-your-own-bot platform (Conversation Builder), not a single fixed AI persona — enterprises assemble bots to their own spec, including small-talk/chitchat behavior LivePerson explicitly offers to make bots feel more human-like. At enterprise scale, the Article 50(1) question isn't whether LivePerson supports disclosure — it clearly can — but whether each brand's custom-built bot actually includes it, since the platform provides the tools without mandating their use.

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 LivePerson

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

• The messaging entry point (web, SMS, Apple Messages for Business, and other configured channels) and the bot's opening turn
• The bot's configured identity and conversational style, including any small-talk/chitchat settings that push it toward sounding human
• The bot's first generated or scripted reply before a live agent is engaged
• The transfer-to-agent moment, which LivePerson can precede with a "Message to User" notification that always displays before the handoff

Your LivePerson disclosure checklist

• Open your site or messaging channel as a visitor: is the AI nature clear in the bot's first message across every channel you've deployed it to (web, SMS, social messaging)?
• Review your bot's build in Conversation Builder for small-talk/chitchat settings — features designed to feel human-like need an explicit AI line to balance them
• Use LivePerson's "Message to User" transfer field to make the bot-to-agent handoff explicit, not just technically logged
• Audit each channel and each bot separately — Conversational Cloud supports many brand-specific bots, and a fix in one doesn't propagate to another
• Check voice/IVR-adjacent bot deployments too, if you use them, since Art. 50(1) covers AI interaction generally, not just text chat
• Screenshot each channel's first-interaction state and export the relevant bot's Conversation Builder configuration, dated

Watch out for

Enterprise platforms like LivePerson put the compliance decision in the hands of whoever builds each bot, not the vendor. That decentralization means a large organization can easily end up with a well-disclosed flagship bot on its main site and a half-dozen quietly undisclosed bots built by different teams for different channels — each individually small, collectively a real gap.

Common questions