Key takeaways
- Most guests have heard of AI and many younger diners are open to AI-assisted ordering, but most Americans want to know when AI is involved.
- Pushback comes from feeling trapped, misled or misunderstood, not from automation itself.
- Disclose in the first sentence, offer a person in the second, and keep both short.
- Measure acceptance at your own restaurant: early hang-ups, requests for a person and complaints.
The short answer
Some guests do mind, many do not, and almost all of them mind being misled. Survey data suggests openness is real, especially among younger diners, but it comes with a clear condition: people want to know when they are dealing with AI, and they want a way out if it is not working. A restaurant that meets those two conditions removes most of the friction.
Restaurant operators feel the uncertainty. Customer acceptance ranks among their top three AI concerns, cited by 41.9% of operators in National Restaurant Association research[1]. The rest of this article looks at what guests say, where they push back, and how to word the start of an AI-answered call.
What the surveys say
95%
of US adults have heard at least a little about AI[2]
76%
say it is extremely or very important to be able to tell whether content was made by AI or by people[2]
Awareness is near universal: Pew Research Center finds 95% of US adults have heard at least a little about AI[2]. Transparency is the strongest expectation: 76% say it is extremely or very important to be able to tell whether pictures, videos and text were made by AI or by people[2]. That finding is about content rather than phone calls, but the principle carries over directly to a voice on your line.
Openness to AI in ordering is substantial among younger adults. National Restaurant Association research found half of Gen Z adults (50%) and millennials (52%) say they would consider ordering from an AI-generated video assistant[3].
Source: National Restaurant Association, 2025[3]
Show the data
| Item | Value |
|---|---|
| Gen Z adults | 50% |
| Millennials | 52% |
Where guests push back
Complaints about automated phone lines tend to follow a few patterns. Each one is a design problem you can fix, rather than a reason to avoid automation altogether.
- Feeling trapped: no clear way to reach a person, or a system that keeps trying after the guest has asked for one.
- Feeling misled: a voice that pretends to be a named human employee, or that dodges the question “am I talking to a person?”.
- Being misunderstood: the agent mishears a name or date and the guest has to repeat it several times.
- Wrong answers: holiday hours or menu details that are out of date.
- The wrong kind of call: an upset guest or an allergy question met with cheerful scripted replies.
- Long introductions: a greeting that takes fifteen seconds before the guest can speak.
Who is most comfortable
Comfort with ordering technology is high among younger guests. In 2024 research, 82% of Gen Z adults said they were comfortable placing an order at a limited-service restaurant with a smartphone app[4], and a majority of full-service customers were comfortable using a smartphone app to place an order (63%) or pay (55%)[5].
Treat those as signs of general comfort with digital service, not proof that every guest wants an AI voice. Your callers are not a survey sample. Regulars who have called the same host for years, older guests and anyone with a complicated request may prefer a person, and the agent should make that easy rather than steer them away.
Disclosure wording that builds trust
The best disclosures are short, early and paired with a choice. Say what the guest is talking to, what it can help with, and how to reach a person. Then let them speak.
| Situation | Suggested wording | Why it works |
|---|---|---|
| Overflow during service | “Thanks for calling Harbor Grill. Our team is with guests, so you've reached our automated assistant. I can help with bookings and questions, or get you to a person.” | Explains why, names the help, offers a person |
| After hours | “Harbor Grill is closed right now. I'm the restaurant's automated assistant and can take a reservation request or answer questions. Our team will follow up in the morning.” | Sets expectations about follow-up |
| All calls, short version | “Hi, this is Harbor Grill's automated assistant. How can I help? You can ask for a person any time.” | Fast for repeat callers |
| When asked directly | “Yes, I'm an automated assistant. Would you like me to connect you with someone on the team?” | Honest, then offers a choice |
| With recording | “This is Harbor Grill's automated assistant, and this call is recorded for accuracy.” | Covers both disclosures in one line |
Never give the agent a human name and backstory that implies it is a staff member. A friendly name is fine if it is clearly presented as the restaurant's assistant.
Principles for the first ten seconds
- Restaurant name first, so callers know they reached the right place.
- Disclosure in the first sentence, not after a menu of options.
- A person offered in plain words, not hidden behind a keyword.
- Recording disclosure if calls are recorded, in the same breath.
- Total greeting short enough that the guest can speak within a few seconds.
- Tone that matches the restaurant: warm for a neighborhood bistro, crisp for a busy counter.
Regulars, older guests and complicated calls
Some callers deserve a different path. Regulars who call every Friday may know your hosts by name and feel a change sharply. Older guests may be fine with technology but prefer a conversation for anything beyond a simple question. Planners of large events want to talk through options, not answer a questionnaire.
- Let regulars reach the host directly: some systems can recognize known numbers and ring the host stand first.
- Keep the language plain and the pace unhurried; avoid jargon such as “say your intent”.
- Route large-party and event calls to the events manager early in the call, with a callback if they are busy.
- Make the human option work at any point, not just at the start.
- Tell your hosts about the change, so they can reassure guests at the door who ask about it.
Bring your team along
Guests often take their cue from staff. A host who rolls their eyes about “the robot” when a guest mentions the phone undoes the work of a careful greeting. Explain to the team why you are using an assistant, which calls it takes, and which ones still come to them. Most hosts welcome fewer interruptions during the rush once they see that complaints, allergies and large parties still reach a person.
Give the team a simple line for guests who ask: “Our assistant picks up when we're busy so nobody waits on hold, and it sends anything tricky straight to us.” Then make sure that is true by checking the handoffs each week.
Measure acceptance at your own restaurant
Track early hang-ups
Count calls that end within the first few seconds of the greeting, before and after the change. A rise suggests the greeting is too long or off-putting.
Track requests for a person
Count how often guests ask for a human and at what point in the call. Early requests may mean the greeting needs work; late ones point to misunderstandings.
Read complaints and reviews
Search new reviews and guest emails for mentions of the phone. Note whether complaints are about the AI itself or about waiting for a callback.
Ask a few guests
At the host stand, ask regulars how the phone worked for them. Short, informal feedback often explains what the numbers cannot.
Compare outcomes
Look at bookings and catering leads captured per week against the period before. Acceptance that does not lead to outcomes is not the goal.
Example: rewriting a greeting
Example: a fictional steakhouse launches with a greeting that lists the specials and the hours before mentioning that the caller has reached an assistant. In the first week, early hang-ups rise and several regulars ask the host whether the restaurant has “gone robot”. The manager cuts the greeting to one sentence with the disclosure and an offer to reach a person.
Example: in the following weeks early hang-ups fall back, requests for a person drop because guests know they can ask, and the host stops hearing complaints at the door. The manager also adds a rule that any caller who says “person” or “human” is transferred or offered a callback at once. See when an AI phone agent should hand off to staff for the full trigger list, and questions to ask before buying an AI phone system to check that disclosure wording is editable.
Frequently asked questions
Do restaurant guests accept AI answering the phone?
Many do, especially when it is disclosed and a person is easy to reach. Most Americans say it matters to know when AI is involved, so honesty in the greeting is the key.
Should we tell callers they are talking to AI?
Yes. Say it in the first sentence, explain what the assistant can help with and offer a person. Hiding it is the fastest way to lose trust.
Can we give the AI assistant a name?
A friendly name is fine if it is clearly presented as the restaurant's automated assistant, not as a human staff member.
How can we tell if guests dislike it?
Track early hang-ups, requests for a person, complaints and booking outcomes before and after launch, and ask a few regulars directly.
References
National Restaurant Association (2026). Research Insight: The Future of Restaurant Technology (press release, November 5, 2026) (opens in a new tab) ↑
Pew Research Center (2025). How Americans View AI and Its Impact on People and Society (opens in a new tab) ↑
National Restaurant Association (2025). From Trend to Transformation: Off-Premises Dining Now Essential for Restaurant Consumers, Operators (opens in a new tab) ↑
National Restaurant Association (2024). Serving Up Technology: New Data Shows How Tech Integration is Transforming the Restaurant Experience (Restaurant Technology Landscape Report 2024) (opens in a new tab) ↑
National Restaurant Association (2024). Restaurant Technology Landscape Report 2024 (opens in a new tab) ↑