Customer Experience & Conversational AI Practical insights
Disclose your AI assistant clearly from the start
Make AI assistant disclosure visible and understandable at first contact, then test comprehension across chat, voice and mobile without burying the message.
A customer should understand when they are interacting with an AI assistant. Clear disclosure helps them interpret the response and choose an appropriate next step. Design it as part of the opening experience, rather than relying on a policy link that few people notice during a conversation.
Understand the applicable obligation
The European Commission’s Article 50 guidance describes a provider obligation to inform people about direct AI interaction unless that is obvious. Notification should be clear, distinguishable and accessible from the first interaction. Article 50 applies from 2 August 2026. [1] Determine the organization’s role and applicable obligations with qualified review; this design workflow is not a complete legal assessment.
Write a short, useful introduction
A proposed opening is: ‘I’m this company’s AI assistant. I can help with product questions and connect you with our team when needed.’ Match the stated abilities to the real service. Use a visible label in chat and an audible introduction in voice. Avoid a human job title or portrait that undermines the disclosure.
Keep the context understandable
A returning visitor may enter a conversation through a link or reopen a widget midway. Check that the AI identity remains apparent without repeatedly interrupting the task. Show how to obtain human help when that option exists. Keep privacy information available separately; an AI label does not by itself explain data use or establish consent.
Test what people understood
Ask test participants who they thought was responding, what the assistant could do and how they would reach another channel. Review mobile layouts, screen-reader output and voice introductions. Measure comprehension alongside task completion rather than optimizing solely for the shortest greeting. If users mistake the assistant for a person, revise the wording, placement or surrounding design and test again. Preserve the results and approved wording as part of the service’s release documentation.
Sources and evidence
Sources checked on 4 October 2026. Proposed workflows and hypothetical examples are editorial analysis.
From insight to practice