Should Your AI Receptionist Tell Callers It Is an AI?
The technology is now good enough that most callers do not notice. That is exactly why the question is worth answering deliberately rather than by default.
This comes up in nearly every onboarding call. The assistant sounds natural enough that most callers will never ask, so the decision falls to the business owner: do you say so, and if you do, when?
We have a view, but it is a view rather than a rule, and the right answer differs by industry.
The three positions
In practice businesses pick one of three postures. Disclose upfront, in the greeting. Disclose on request, where the assistant answers honestly the moment anyone asks. Or say nothing at all.
We will not configure the third. Every assistant we build answers truthfully when a caller asks whether they are talking to a person. Beyond that, the choice is yours.
The case for disclosing upfront
- It removes the worst outcome entirely. A caller who works it out mid-conversation and feels deceived is far more damaging than one who knew from the start.
- It changes how people talk, usually helpfully. Callers who know tend to be more direct, which produces cleaner bookings and fewer misunderstandings.
- In regulated or sensitive contexts, it is increasingly the expectation rather than a nicety.
- It is a differentiator. Being the practice that is straightforward about it reads as confidence, not apology.
The case against
- A minority of callers hear the word automated and hang up before discovering the assistant could have booked them in thirty seconds.
- It front-loads the greeting with information the caller did not ask for when they just want your opening hours.
- If the interaction is going to be brief and transactional, the disclosure can be more disruptive than the fact.
Where we land
For most businesses, a light upfront disclosure in the greeting is the right call. Not an announcement, just a phrase. Something like: thanks for calling Riverside Dental, this is Jane, the practice's automated assistant, how can I help? It takes under two seconds, it sets expectations, and it removes any chance of the caller feeling misled later.
For healthcare, legal, and financial services we would go further and treat upfront disclosure as the default, because the sensitivity of what callers are about to say makes informed consent the point rather than a courtesy.
The legal layer, briefly
Disclosure of AI and consent to record are two different obligations and people conflate them. Several US states require all-party consent to record a call, which means an announcement regardless of whether a human or an assistant is answering. Some jurisdictions have added, or are adding, specific requirements around disclosing automated agents. This varies by where you and your callers are, and it is worth ten minutes with your own counsel rather than a paragraph from your vendor. What we can tell you is what the assistant is capable of announcing, and the answer is anything you want it to.
What actually determines the reaction
In our experience, whether a caller minds has very little to do with the disclosure and almost everything to do with competence. People do not resent talking to an assistant. They resent not getting their problem solved. An assistant that books the appointment, answers the question, and transfers cleanly when it should gets thanked. One that loops, mishears, and dead-ends gets complained about, and the complaint will mention that it was a robot, but that is not really why.
Callers forgive a machine that helps them. They do not forgive one that wastes their time, and they never forgive one that lied about what it was.