Concierge legal

AI Disclosure

Concierge uses artificial intelligence to help website visitors understand approved business information and reach an appropriate next step.

Last updated: August 19, 2026

Visitors should know when they are interacting with AI. The Concierge Agent should disclose its automated nature at the beginning of the interaction unless it is already obvious from the surrounding experience.

Where Concierge uses AI

Generating and refining chat or voice responses from customer-authorized sources and page context.
Classifying intent, urgency, routing needs, and conversation state.
Summarizing conversations and helping authorized teammates understand customer context.
Analyzing public website content, business information, structured data, and growth opportunities.
Helping customers create or organize approved Agent instructions, Pages, and knowledge content.

What AI does not guarantee

AI output can be inaccurate, incomplete, outdated, biased, or inconsistent. It may misunderstand a visitor, source, page, accent, or context. Concierge does not guarantee that an AI response is correct, legally compliant, suitable for a particular person, or sufficient to complete a transaction.

High-stakes and professional decisions

Concierge is not an emergency service and should not diagnose, prescribe, determine legal rights, make credit or insurance decisions, decide employment or housing eligibility, determine child safety, or replace qualified professional judgment. Customers must configure clear boundaries and human escalation for medical, legal, financial, insurance, safety, accessibility, and other consequential matters.

Human escalation and review

A visitor may be offered a human connection, call request, ticket, or another customer-configured route. Availability and response time depend on the relevant business. Authorized customer users may review, join, reply to, assign, resolve, and follow up on conversations.

Customer responsibilities

Use accurate, authorized, and current source material.
Test realistic questions before publishing and after material content changes.
Identify topics the Agent may answer and topics it must route to a person.
Provide legally required notices, consents, and industry-specific controls.
Monitor results and correct sources or configuration when an answer is unsuitable.

Information used by AI providers

Enabled AI features may send relevant instructions, page context, knowledge excerpts, conversation content, or voice-derived text to contracted AI providers to generate the requested result. We limit the information to what is reasonably needed. Details appear in our Privacy Policy and Subprocessors page.

Feedback, complaints, and correction

Visitors should use the business's human contact path for an incorrect or consequential answer. Concierge customers can update sources, instructions, and routing. Privacy requests can be submitted through Privacy Choices; suspected misuse can be reported through Report Abuse.