🇺🇸 United States · Intercom Fin — AI Customer Service Agent
Status: 🟩 COMPLETE 🟦 LIVING Section: 10 — AI and LLMs
| Vendor | Intercom |
| Country/origin | 🇺🇸 United States (San Francisco; founded Ireland 🇮🇪) |
| Recommended for AUS? | ✅ Yes — well-established; enterprise data handling; used widely by Australian businesses |
| Privacy summary | SOC 2 Type II; ISO 27001; GDPR; data residency options; enterprise DPA; customer conversation data handled securely |
| Free tier | 14-day trial; no permanent free tier for full Fin AI |
| Paid tiers | Fin AI Agent: typically included in Intercom plans (~132 USD/month base + per resolution fees) |
| First released | Intercom founded 2011; Fin AI agent launched May 2023 |
| Last reviewed | June 2026 |
| Official site | https://intercom.com/fin |
What it is
Intercom Fin is an AI-powered customer service agent — a chatbot that actually resolves customer queries rather than just routing them. It’s built into Intercom’s customer messaging platform and powers the chat widget you see on websites and apps.
“Fin” (short for “final AI”) can:
- Answer customer questions by reading your product documentation, help articles, and website content
- Resolve common queries end-to-end: “How do I reset my password?”, “What is your refund policy?”, “How do I cancel my subscription?”
- Hand off to a human agent when the query is complex or when the customer requests it
- Work 24/7 across multiple languages
- Use conversation history to maintain context throughout a chat
- Be customised with your brand tone and specific instructions
By mid-2026, Fin 2 uses a multi-model AI architecture and significantly improved reasoning capabilities, including the ability to take actions (like looking up order status or triggering refunds) not just answer questions.
How it works (plain English)
Fin reads your company’s knowledge base — all your help articles, FAQ pages, product documentation, and website content. When a customer asks a question, Fin:
- Searches your knowledge base for relevant content
- Synthesises a clear, accurate answer based on what it found
- If it can’t find a confident answer, it says so and offers to connect the customer with a human
This is Retrieval-Augmented Generation (RAG) in action — see rag. Fin grounds its answers in your actual content rather than making things up.
How to set up + first steps (Australian businesses)
- Go to https://intercom.com → Start a trial
- Connect your website chat widget or app messaging
- Add your help centre content: Settings → Help Center → create and publish articles
- Enable Fin: Settings → AI → Fin → toggle on → Fin immediately starts answering based on your articles
- Configure: set which topics Fin handles, what it should escalate, the brand tone
- Monitor: Reports → Fin → see resolution rates, what questions Fin couldn’t answer (opportunities to write more help articles)
The key to Fin performance: the more comprehensive your help centre content, the better Fin performs.
What it costs
Intercom’s pricing is complex and has evolved significantly. As of mid-2026:
- Fin AI is included in Intercom plans, with per-resolution pricing for AI-resolved conversations
- Typical per-resolution cost: ~$0.99 USD per AI-resolved conversation
- Plan costs start at ~74 USD/month (Starter) up to higher enterprise tiers
- For businesses with high support volume, the per-resolution model can become significant — model your expected volume before committing
The value proposition: if Fin resolves 60% of queries that would otherwise need a human agent, and a human agent costs 2.50/query), Fin at $0.99/query is significantly cheaper.
How it compares to alternatives
| Tool | Country | Best for |
|---|---|---|
| Intercom Fin | 🇺🇸🇮🇪 | Best-in-class AI; established platform; businesses already on Intercom |
| Zendesk AI | 🇺🇸 | Zendesk users; large support organisations |
| Freshdesk AI (Freddy) | 🇮🇳🇺🇸 | Mid-market; Freshworks ecosystem |
| Tidio | 🇵🇱🇺🇸 | Small business; affordable |
| Drift | 🇺🇸 | B2B sales-focused chat |
| Sierra AI | 🇺🇸 | Enterprise; voice AI for customer support |
| HubSpot AI chat | 🇺🇸 | HubSpot ecosystem users |
| Crisp | 🇫🇷 | Small teams; EU privacy |
Intercom Fin is widely considered the market-leading AI customer service agent, particularly for SaaS companies. Its resolution rates and quality consistently outperform alternatives in independent comparisons.
Privacy / data handling
- Customer conversation data (which includes personal information) is processed by Intercom and the AI models powering Fin
- SOC 2 Type II and ISO 27001 certified
- GDPR compliant; DPA available
- Australian Privacy Act 1988 applies — ensure your website privacy policy discloses the use of AI in customer support and how conversation data is handled
- Customer consent for AI interaction: Intercom recommends a disclosure in the chat widget (“You’re chatting with Fin, our AI assistant. Your conversation may be reviewed to improve our service.”)
Gotchas
- Fin is only as good as your knowledge base. If your help articles are outdated, poorly written, or sparse, Fin will give poor answers. The AI amplifies the quality of your existing content.
- Per-resolution pricing can surprise. Budget carefully based on your support volume. High-traffic businesses should model this carefully before committing.
- Escalation design matters. Fin needs a clear handoff to human agents. Design your escalation paths thoughtfully — a frustrated customer who can’t reach a human is a bad experience.
- Not suitable for complex, high-stakes support. Medical advice, legal guidance, financial transactions — Fin should quickly escalate these to qualified humans. Configure escalation triggers carefully.
- Multilingual quality varies. English performance is excellent; other languages depend on the language coverage of your help content and the underlying model.
- Australian consumer law note: If Fin is answering questions about returns, refunds, or consumer rights, ensure the answers reflect Australian Consumer Law (not US refund policies). Fin answers based on what you write — you’re responsible for correct, legally compliant content.
See also
- hubspot-ai — HubSpot’s AI including customer service features
- salesforce-einstein — Salesforce service cloud AI
- rag — the underlying technology (retrieval from your knowledge base)
- real-time-voice-ai — AI voice agents for phone support (the voice equivalent)
Sources
- Intercom Fin documentation: intercom.com/fin
- Intercom Fin 2 announcement (2024)
- Intercom pricing: intercom.com/pricing
- Australian Privacy Act 1988 — AI in customer service obligations
- Australian Consumer Law — returns and refund obligations (ACCC)
- TechCrunch and CX industry coverage of Fin (2023–2026)