Intercom Fin

Overview

Intercom Fin is Intercom's AI-powered customer support agent that handles support conversations automatically through the familiar Intercom chat widget. Unlike traditional chatbots that follow rigid decision trees, Fin uses large language models trained on a business's help center content, documentation, and past support conversations to provide contextual, conversational answers to customer questions. It operates within Intercom's existing messenger framework, meaning websites already running Intercom's chat widget can activate Fin as the first-line support responder.

For website operators, Intercom Fin's script presence is identical to the standard Intercom Messenger — it is the same JavaScript widget, enhanced with AI response capabilities on the backend. The consent considerations center on the data the widget collects and how it identifies visitors, rather than on the AI processing itself.

What This Script Does

Intercom Fin's scripts embed a chat interface and perform several data collection and communication functions:

• Chat widget rendering: Displays the Intercom Messenger interface, typically as a floating button that expands into a conversation panel. The widget adapts to mobile and desktop layouts.
• AI-powered responses: Routes incoming messages to Fin's AI engine, which generates responses based on the business's knowledge base. Fin can answer questions, provide links to relevant help articles, and escalate to human agents when it cannot resolve an issue.
• Visitor identification: Sets cookies to identify returning visitors and maintain conversation continuity across sessions. If the visitor is a logged-in user, the widget associates the conversation with their account profile in Intercom.
• Conversation history: Stores and retrieves past conversation threads so returning visitors can continue previous support interactions without repeating context.
• User data collection: Captures information provided during conversations, including name, email address, and any details shared in the chat. This data is stored in Intercom's contact database.
• Behavioral context: Collects the current page URL and navigation path to give support agents (and the AI) context about where the visitor is and what they might need help with.
• Proactive messaging: Can display targeted messages, product tours, or announcements based on visitor behavior, segment membership, or page-specific triggers configured by the site operator.

Consent & Compliance

Intercom Fin's chat widget serves a primarily functional purpose — providing customer support — but its data collection practices extend beyond what is strictly necessary for a single chat exchange. The persistent identification cookies, behavioral tracking, and proactive messaging features introduce consent considerations.

Key compliance factors:

• Functional core: The chat widget itself serves a legitimate support function. Visitors who initiate a conversation are explicitly requesting a service, and the data collected during that conversation supports service delivery.
• Persistent identification: Cookies that identify returning visitors across sessions go beyond the immediate chat function. Under strict interpretations of ePrivacy rules, these may require consent even though they improve the user experience.
• Proactive messaging: If configured to display targeted messages based on behavior or segments, these features blur the line between support and marketing, potentially requiring marketing consent.
• Data storage: All conversation data is stored in Intercom's platform. Ensure a DPA is in place and that data subject access and deletion requests can be fulfilled through Intercom's tools.

The consent classification depends on how the widget is configured. A basic support chat is functional. A fully configured Intercom installation with proactive campaigns, behavioral targeting, and user tracking moves into territory that warrants consent.

Should You Block This Without Consent?

Intercom Fin serves a functional customer support purpose, and blocking it would prevent visitors from accessing live help. However, its persistent identification cookies and optional proactive messaging features extend beyond strictly necessary functionality. If used purely for reactive support chat, it can reasonably be loaded without consent. If proactive messaging or behavioral targeting features are enabled, those components warrant consent. No.