As AI becomes part of everyday work, a fair question comes up: when should you tell people AI was involved? There is no single rule for every situation, and laws vary by place and industry. But a few principles keep you honest and protect the trust you've built with customers and colleagues.
When to tell customers
You don't need to label every email that AI helped you tidy up. If you wrote the content, checked it, and stand behind it, AI was a tool, much like spell-check.
Tell people when AI use would matter to them:
- They are talking to an AI, not a person. Website chatbots, AI voice agents on your phone line, and automated text replies.
- AI shaped something they rely on. A report, an estimate, or advice where they would reasonably expect a person did the work.
- Their information is handled by AI. For example, if call recordings or intake forms are summarized by an AI tool. Your privacy notice should say so, and some laws may require it. Check with a qualified professional.
- They ask. If a customer asks, "Did a person write this?" answer honestly.
A simple test: if the customer found out later, would they feel misled? If yes, tell them up front.
Never pretend an AI is a person
This is the clearest line. If a customer is chatting with a bot or talking to an AI voice agent, they should know it. Don't give a chatbot a human name and photo and let customers believe it's a staff member. Don't let an AI voice agent introduce itself as someone from your front desk.
Good practice for any AI that talks to customers:
- Say it's an AI at the start: "Hi, I'm the virtual assistant for [Business Name]."
- Make it easy to reach a person: "Say 'staff' at any time to talk to someone on our team."
- Don't let it make promises or decisions it shouldn't, such as refunds, medical advice or legal answers.
Being upfront doesn't spoil the experience. Many customers are happy to use a helpful assistant. What upsets people is finding out they were fooled.
Telling colleagues
Inside your business, be open too. If you used AI to draft a report, a proposal or an SOP, say so when you pass it on: "First draft made with AI. I checked the numbers. Please review section 3." That tells the reader where to look closely and keeps everyone clear about where content came from.
Never pass off AI work as checked if you didn't check it.
Keep a human in charge of decisions about people
AI can help sort, summarize and draft. It should not make the final call on decisions that affect people's lives or livelihoods, such as:
- Hiring, firing, promotions or pay
- Accepting or turning down a tenant application
- Offering credit or payment plans
- Patient care or treatment
In these cases a person reviews the information, makes the decision, and can explain why. If AI helped, that person checks its work and is ready to set it aside. Some laws place rules on automated decisions like these, so check with a qualified professional before using AI anywhere near them.
Can this go into AI? A property manager wants to paste a stack of rental applications, with names, incomes and credit details, into an AI tool and ask it to pick the best tenant.
Answer: No — the applications are full of personal financial data, and choosing a tenant is a decision about a person that a human must make. Review applications against your written criteria instead.
Can this go into AI? A front-desk employee wants AI to draft the greeting for the office's new phone voice agent.
Answer: Yes — it's your own business wording, with no personal data. Make sure the greeting says plainly that callers are speaking with an automated assistant and tells them how to reach a person.
Key points
- Tell customers when they are dealing with an AI, when AI shaped something they rely on, or when they ask.
- Never let a chatbot or voice agent pretend to be a person, and always offer a way to reach a human.
- Be open with colleagues about AI drafts, and never claim you checked something you didn't.
- People, not AI, make decisions about hiring, tenants, credit and care.
AI disclosure lines
Ready-made wording for telling customers and colleagues when AI is involved.
- Chatbot greeting: Hi, I'm the virtual assistant for [Business Name]. I can help with [topics]. Type 'staff' at any time to reach a person.
- Voice agent greeting: Thanks for calling [Business Name]. You've reached our automated assistant. Say 'staff' or press [number] to leave a message for our team.
- Privacy notice line: We use AI tools to help [summarize calls / draft replies]. A member of our team reviews [what is reviewed]. Questions? Contact [email].
- Honest answer to 'Did a person write this?': Our team used an AI tool to help draft this, and [your name] reviewed it before sending.
- Note to a colleague: First draft made with AI. I checked [what you checked]. Please review [section or detail].