Should You Tell Customers They Are Talking to an AI?
We found no ICO guidance that specifically requires you to tell callers they are speaking to an AI, but telling them up front is good practice. It builds trust, avoids awkward moments and fits the transparency principles the ICO promotes. Rules in other countries differ, so check where your callers are.
What does the ICO say about telling people it is an AI?
The ICO is the UK data protection regulator. Its guidance on AI and data protection explains that the right to be informed under UK GDPR covers the existence of solely automated decisions that have legal or similarly significant effects. It also covers meaningful information about the logic involved in those decisions.
That is about automated decisions, not about every conversation with a chatbot or voice assistant. We checked the ICO's legal framework page for this guide and it does not say that you must announce to callers that they are speaking to an AI. We did not find ICO guidance that specifically requires this, and we are not going to invent a rule that is not there.
The ICO and the Alan Turing Institute also published guidance called Explaining decisions made with AI. Its principles are to be transparent, be accountable, consider the context and reflect on the impact. Those principles are about how you explain AI-assisted decisions, but they point in a clear direction: be open about how you use AI.
The ICO has also said it will develop a statutory code of practice on AI and automated decision-making. Treat that as planned. Check the ICO website for its current status before relying on anything you read about it, including here.
Does the law in other countries require it?
Some countries and regions have their own rules on telling people when they are interacting with an AI system, and the EU is one example. We have not set out those rules here, because they differ by place, they change, and the wording matters. If you have callers or customers outside the UK, check the rules for each place or ask a qualified adviser.
Our guide to AI call answering and recording rules by country is the place to start for the legal side, and the guide on whether it is legal to use AI to answer business calls in the UK covers the UK position.
Why say it up front even if the law may not demand it?
The business case is plain and does not depend on a specific rule.
Trust. People who find out later that they were talking to a machine can feel misled, even if the call went well. Saying it first removes that risk.
Fewer awkward moments. A caller who knows it is an AI will phrase things simply and will not be thrown by a small pause or an odd reply.
Better calls. Callers who know what they are talking to tend to say what they want plainly, which helps the assistant do its job.
Consistency. If your privacy notice and website say you use AI, your phone line should not pretend otherwise. That matches the transparency principles above.
Less to defend. If a complaint or a regulator's question ever arises, a clear opening line is easy to point to.
When does telling people matter more?
Some situations raise the stakes, and in these a clear disclosure is close to essential rather than just polite.
I
n
t
h
e
s
e
c
a
s
e
s
,
a
l
s
o
d
e
c
i
d
e
w
h
e
n
t
h
e
a
s
s
i
s
t
a
n
t
s
h
o
u
l
d
s
t
o
p
a
n
d
p
a
s
s
t
h
e
c
a
l
l
t
o
a
p
e
r
s
o
n
.
O
u
r
g
u
i
d
e
o
n
w
h
e
n
A
I
s
h
o
u
l
d
h
a
n
d
o
v
e
r
t
o
a
p
e
r
s
o
n
c
o
v
e
r
s
t
h
a
t
.
Sensitive topics, such as health, money worries, legal matters or bereavement.
Anything that feeds into a decision that affects the person, such as whether they get a service, a booking, a refund or a quote. This is where the ICO's points about automated decisions and meaningful information may apply, so take advice.
Calls that are recorded or transcribed, because personal data is being collected and people should know how.
Vulnerable callers, who may be more easily confused or distressed by not knowing who or what they are speaking to.
How do you say it? Example opening lines
Keep it short, friendly and in plain words. The disclosure should take a few seconds, not read like a legal notice. These are examples to adapt to your own business, not required wording.
T
h
e
b
u
s
i
n
e
s
s
n
a
m
e
s
a
b
o
v
e
a
r
e
m
a
d
e
u
p
f
o
r
i
l
l
u
s
t
r
a
t
i
o
n
.
U
s
e
y
o
u
r
o
w
n
n
a
m
e
a
n
d
d
e
s
c
r
i
b
e
o
n
l
y
w
h
a
t
y
o
u
r
a
s
s
i
s
t
a
n
t
c
a
n
r
e
a
l
l
y
d
o
.
What if someone asks, "Am I talking to a robot?"
Always answer truthfully. Set the assistant up so that a direct question about whether it is a person or an AI gets a direct, honest yes, never a dodge or a joke that leaves the caller unsure.
Then offer a way forward: a person to speak to, a call back, or the option to carry on with the assistant. A reply such as "Yes, I'm an AI assistant. Would you like me to arrange for someone to call you back, or can I help here?" works well.
Test this before you go live. Call your own line, ask the question in a few different ways, and check the answer every time. Our guide on whether an AI voice assistant can make things up explains why you should test rather than assume.
What about recording and personal data?
Telling people it is an AI is a separate point from telling them how their data is used. If calls are recorded, transcribed or stored, callers should be told, and your privacy notice should explain what you collect, why and for how long.
We keep the legal detail in other guides so it stays in one place. See is it legal to use AI to answer business calls in the UK and AI call answering and recording rules by country. For the basics of what the technology does, see what is an AI voice assistant.
A short checklist before you go live
The opening line says it is an AI, in plain words.
The assistant answers honestly if asked whether it is a person.
A person is always on offer, and the hand-over works when you test it.
Recording and data use are mentioned on the call where needed and covered in your privacy notice.
You have checked the rules for every country your callers are in.
Your website and privacy notice match what the phone line says.
You have checked the ICO website for the current status of its planned AI code of practice.
Sources
ICO: legal framework for explaining decisions made with AI
https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/explaining-decisions-made-with-artificial-intelligence/part-1-the-basics-of-explaining-ai/legal-framework/
ICO: artificial intelligence and biometrics strategy, plan of action
https://ico.org.uk/about-the-ico/our-information/our-strategies-and-plans/artificial-intelligence-and-biometrics-strategy/our-plan-of-action/
Erudience services related to this guide
Terms used in this guide
More guides to read next
- Is It Legal to Use AI to Answer Business Calls in the UK?
- AI Call Answering and Call Recording Rules: A Global Overview
- When Should AI Hand a Customer Over to a Person?
- What Is an AI Voice Assistant? A Plain Guide for Business Owners
- Can an AI Voice Assistant Make Things Up? How to Keep It Honest
Frequently asked questions
Do I have to tell callers they are speaking to an AI in the UK?
We found no ICO guidance that specifically requires it. The ICO's rules on the right to be informed cover solely automated decisions with legal or similarly significant effects, which is a narrower point. Telling callers is still good practice, and you should take advice if the assistant influences decisions about people.
What if my callers are in the EU or another country?
Other countries and regions have their own rules on telling people when they are interacting with an AI system, and the EU is one example. They differ and change, so check the current rules for each place or ask a qualified adviser rather than relying on a general guide.
Will telling callers it is an AI make them hang up?
It may put some people off, and we have no figures on how often. Others are reassured by the honesty. A short, friendly line that offers a person if they prefer is the safest approach. Track your own call outcomes after you launch to see how callers react.
What should the assistant say if a caller asks whether it is a robot?
It should say truthfully that it is an AI assistant, then offer to arrange a person or carry on helping. It should never claim to be human or avoid the question. Test this on your own line before launch.
Is telling callers it is an AI the same as telling them the call is recorded?
No. They are separate points. Disclosing the AI is about transparency, while recording and storing calls involves personal data and has its own notice requirements. Our guides on UK legality and recording rules by country cover the legal side.
Is the ICO going to publish a code of practice on AI?
The ICO has said it will develop a statutory code of practice on AI and automated decision-making. That is planned, so check the ICO website for its current status before you rely on it.