Telling Customers When They're Talking to AI

So you've added an AI chatbot or agent to your website, inbox, or phone line. Should you tell customers it's AI? Our take: yes, kindly and clearly. People feel better when they know who, or what, they're talking to. And trust is your small business's superpower.

The good news is it's easy to do well. Here's how.

First, say it up front. A short line at the start of the chat is enough, like "Hi! I'm the virtual assistant for [business]. I can help with most questions, and a person is here if you need one." Next, make a person easy to reach. Tidio says you can set handoff rules and watch chats live to jump in anytime, and Fin says it can transfer conversations to your team's inbox. Then, sign AI-drafted emails honestly. If a person reviews and sends the email, it's from you. If something is sent automatically, say so in a friendly way. After that, keep it on your own info. Zapier Chatbots, Tidio, and Fin all describe training on your own FAQs, docs, or help center, which keeps answers grounded. And last, check the rules. Some places have rules about AI disclosure, recorded calls, or automated messages. Check what applies to you.

Wording you can copy:

• For website chat: "I'm [business]'s AI assistant. Ask me anything, or type 'person' to reach our team."
• For a hand-off: "Great question. I'm connecting you with [name] from our team now."
• For automated emails: "This reminder was sent automatically. Just reply if you need a hand."
• For phone or voice: "You're speaking with our virtual assistant. Say 'team member' at any time."

A few things to avoid:

• Giving the bot a human name and photo: it can feel misleading. A friendly bot name is fine, as long as it's clear it's AI.
• Hiding the way out: never make people fight to reach a person.
• Pretending it's perfect: a simple "I might not get everything right" sets fair expectations.

Honesty doesn't make AI less helpful. It makes your customers more comfortable using it.