AI in Business Operations

How Should a Small Business Use AI for Customer Service?

AI belongs in small business customer service, but not where most owners start. Here is what customers want, who is liable when a chatbot gets it wrong, and the Promise Rule for deciding what AI may say to a customer.

An orange cut-paper service bell on a hand-drawn black counter, standing for the moment a customer asks for a person rather than a machine.

In brief

A small business should use AI in customer service to answer questions it has already answered in writing and to draft replies that a person checks and sends. AI should never promise anything a person has not decided, and upset customers, money outside policy and new problems go straight to a person. Start in the inbox, not with a website chatbot.

A small business should use AI in customer service to draft replies and to answer questions it has already answered in writing, with a person approving anything that makes a promise, involves an upset customer or touches money. Start in the inbox, not with a website chatbot. Customers want a route to a human, and a tribunal has already held a company to what its chatbot told a customer.

Sources re-read November 20, 2026. This is practical guidance, not legal advice.

Picture a Saturday night in November. A customer emails to ask whether the patio heater she ordered will arrive before her party on Friday. Nobody will read the email until Monday. That is the moment most owners have in mind when they ask whether AI should handle customer service, and it is why the answer below is yes, but not in the job most of them first imagine.

Where does AI actually help in small business customer service?

AI helps most with the reading and writing in customer service, and least with the deciding.

Look at what fills a customer inbox in a typical small business. Most of it is reading: working out what the customer wants, which order they mean, whether this is the third message about the same problem. Then comes writing: a clear, polite reply that gets the facts right. Deciding is a thin slice at the end, and it is the slice that matters most.

AI is good at the first two. Inside the tools many businesses already pay for, such as Microsoft Copilot in Outlook, Gemini in Gmail or ChatGPT, a member of staff can:

  • sort the morning's messages into orders, returns, complaints and questions;
  • summarize a long email thread before replying to it;
  • draft a reply from your own written policies, in your house tone;
  • turn a scrawled phone note into a tidy order update;
  • rewrite a help page so a customer can find the answer before they need to ask.

None of these puts AI in front of a customer on its own, which is deliberate.

Most small firms are still near the beginning. The UK's Office for National Statistics reported on July 20, 2026 that around 35% of businesses with ten or more employees now use at least one AI technology, and 28% of businesses with 0 to 9 employees, against 49% of those with 250 or more (Business Insights and Conditions Survey, fieldwork 5 to 28 June 2026). It also found that improving business operations is the most reported use of AI in every size band. Customer service sits right on that line: an operation behind the scenes, with a customer at the end of it.

What do customers want from AI customer service?

Customers welcome AI that makes contact easier and resent AI that stands between them and a person.

A Gartner survey of 3,566 business and consumer customers, conducted in February and March 2026 and published on August 4, 2026, found that 50% of customers say their interactions are easier when companies use generative AI. It also found that 87% say it is essential for a company using it to offer a way to reach a human agent. Eric Keller, a senior director analyst at Gartner, put the conclusion in one line: "Service leaders should not use GenAI as a mandatory first step for every issue."

Two further findings from the same survey matter more to a small business than the headline. In their most recent service interaction, customers were about three times more likely to use their own AI tools, such as ChatGPT, Gemini or Copilot, than a company's chatbot. And 58% of customers who use generative AI have used it to complete a task on their behalf, rising to 74% in business-to-business settings.

My reading: your customers are already bringing AI to you. Some will have asked ChatGPT about your returns policy before they email. Some of the emails you receive were drafted by AI. That changes the first job. Before you build a chatbot, make sure the answers it would give are written down clearly on your own site, because machines as well as people are now reading them.

Who is responsible when an AI chatbot gets it wrong?

Your business is, in the same way it is responsible for anything else on its website.

The case owners should know is Moffatt v. Air Canada. A customer asked the airline's website chatbot about bereavement fares and relied on its wrong answer. On February 14, 2024, British Columbia's Civil Resolution Tribunal ordered the airline to compensate him, awarding approximately C$650 in damages plus interest and fees. Air Canada had argued that the chatbot was a separate legal entity responsible for its own actions. The tribunal rejected that and held the airline responsible for all the information on its website, whether it came from a static page or a chatbot.

It is a tribunal decision, not a court judgment and not a rule in the UK or the US, and the sum involved was small. But the reasoning is ordinary common sense, and it is the reasoning I would plan around. A chatbot does not have its own opinions about your refund policy. It has yours, or it has something it made up that customers will treat as yours.

The Promise Rule: what AI may say to your customers

The Promise Rule is my suggested way of deciding what AI may say to a customer, and it sorts by promise rather than by topic.

The Promise Rule is the line I suggest a small business draws before AI speaks to a single customer: AI may tell a customer anything the business has already written down, but it may never promise anything a person has not already decided. Questions with a written answer can go to AI. Replies that make a promise are drafted by AI and sent by a person. Anything involving an upset customer, money outside policy or a problem nobody has seen before goes straight to a person.

Here is how it sorts a week of messages in a hypothetical 15-person online garden supplies retailer heading into its busiest quarter.

Message Written answer exists? Makes a promise? Lane
"What time does your phone line open?" Yes No AI can answer
"Where is my order?" Yes, the tracking record No AI can answer
"Can I return this after 40 days?" (policy says 30) No, it needs an exception Yes AI drafts, a person decides and sends
"Will it arrive by Friday?" Partly Yes, a delivery date AI drafts, a person sends
"Second broken planter this month. Really disappointed." No Whatever you say next is one A person, from the first word

Go back to the patio heater. Under the Promise Rule, AI can tell the customer on Saturday night what the tracking record says, because that is written down. It cannot tell her the heater will arrive by Friday, because nobody has decided that. The honest reply says what the record shows and when a person will confirm. It is less satisfying than "Yes, no problem!" and far cheaper than a promise you cannot keep.

Notice what the rule does not ask. It does not ask whether the AI is clever enough. Modern tools can write a fluent answer to every row in that table. The rule asks whether a person in your business has already made the decision the answer contains. The same thinking applies before a customer ever writes to you: the Receipt Rule keeps AI-drafted marketing and sales messages to claims the business can back.

The Mistake I See Most Often

The mistake I see most often is starting with the chatbot on the website, the most visible and riskiest place, instead of the inbox behind it, where a person still reads every word.

When I work with business teams on customer service, the conversation almost always opens with the widget in the corner of the homepage. It is what the owner has seen elsewhere and what vendors show. But the website chatbot is the one place where AI talks to customers with nobody watching. The inbox is the opposite: AI drafts, a person reads, the customer only sees what the person approves. Every mistake is caught, and every draft teaches the team something about what good looks like.

That last point comes from my first career. I spent two decades as a teacher, assistant headteacher and Director of Digital Strategy in further education before I worked with businesses, and the most reliable way I know to raise the quality of anyone's writing is to let them edit good examples. An AI draft that a skilled customer service person corrects every morning is a training program in disguise. As I put it on my business page, the hard part of AI at work isn't the technology. It's helping people use it well, safely and every day.

The principle underneath is the one I teach everywhere: outsource the doing, not the thinking. Drafting a reply is doing. Deciding what your business will promise is thinking.

A copyable one-page AI customer service sheet

This is the sheet to put in front of whoever answers your inbox before any AI drafts a word. Copy it, fill in the brackets and keep it where the team can see it.

AI customer service sheet for [business name], agreed by [owner] on [date]

  1. AI may answer, from our written answers only: [opening hours], [delivery times and tracking], [returns policy as written], [how to contact a person].
  2. AI may draft, a named person reads and sends: any reply that gives a date, a refund, an exception, a price or a promise. Drafts are never sent unedited.
  3. A person takes it from the first word: an upset or distressed customer, any complaint that mentions a safety, legal or regulatory issue, money outside policy, a problem we have not seen before, and any customer who asks for a person.
  4. AI never sees: card numbers, passwords, health details or anything a customer has asked us to keep private.
  5. Every AI-drafted reply that goes out is logged with who approved it. The owner reads [10] a week and writes down every error.
  6. Review date: [date, four weeks from today]. We widen the first lane only if the error log is clean.

Four weeks of that error log is also your evidence for the Exit Question below. If the log shows a drafted reply that would have made a promise nobody had decided, the rule has done its job.

What should a small business do this week?

This week, sort your real messages, write down your answers and start AI in the inbox before it goes anywhere near a website. A 20-person business can do this in a few hours.

  1. Pull your last 50 customer messages. Emails, web forms, social messages, whatever you have. Sort each into the three lanes of the Promise Rule. Most owners are surprised how many sit in the first lane.
  2. Write down the answers to your ten most common first-lane questions in one short document. This becomes your single source of truth. If an answer is not written down, AI does not say it. To test whether those ten answers are ready for AI to read, use the Single Source Test in Is your business data ready for AI?.
  3. Give AI drafting to the person who already answers the inbox. For two weeks, they draft second-lane replies with the AI tool you already pay for and send nothing unedited. Keep payment details and anything sensitive out of the prompt, in line with your one-page AI policy.
  4. Add one line to that policy: "AI may not promise a customer anything a person has not already decided."
  5. Put the Exit Question to the trial before it starts. The Exit Question is the one question I put to a leadership team before any AI pilot starts: if this works, what will change, who has already agreed to it, and on what date will you decide? A pilot without an answer is not a pilot. It is an experiment with no exit.
  6. Only then consider a customer-facing chatbot, limited to first-lane questions, fed only from your written answers, with a visible route to a person on every screen.

If this feels slow, it is the same logic as starting with one job rather than one tool. Customer service is simply the job where the customer sees your mistakes.

When should a customer message go straight to a person?

Route the message to a person, whatever the written answers say, when it matches any row of this table.

Signal in the message Why it goes to a person What the person does first
Anger, distress or a repeat complaint The customer is judging the business, not the answer Reads the whole thread and replies in their own words
A request for a refund, discount or exception It is a decision about money Checks the policy and decides
A date or delivery promise Nobody has agreed it Confirms with the person who controls stock or the courier
A safety, legal or regulatory mention A wrong reply carries real risk Takes advice before replying
"Can I speak to someone?" Gartner found 87% of customers want that route to exist Replies the same day if possible

When should AI stay out of customer service?

Keep AI away from your customers entirely if your answers are not yet written down, if a handful of large clients make up most of your revenue, or if customers come to you for regulated advice.

The first case is the most common. A chatbot with nothing to read will improvise, and improvisation is exactly what the Promise Rule forbids. Write first. The second case is a business-to-business firm where every relationship is personal and each account is worth a year of effort: there, AI belongs behind the scenes, drafting and summarizing, never answering. The third covers financial, legal, medical and similar advice, where a wrong answer is more than an apology. Charities and social enterprises that serve people in difficulty face a sharper version of the same question, and AI for small charities sets out the Distance Rule for it.

The counterargument deserves its due. A one-person business drowning in the same five questions has the most to gain from answering them automatically and the least time to wait. For that owner, a small first-lane chatbot, fed from a clear written page, may be the right first move. The rule still holds: it answers what is written, and it hands everything else to you. A chatbot that answers is also a long way from software that acts for you: letting AI book, send or pay on the business's behalf is a separate decision, and I set out the test for it in what AI agents mean for small businesses.

Owners who skip these steps often end up with AI that everyone pays for and nobody trusts, the pattern in why most small businesses are not getting value from AI yet.

Where to go from here

If your team already uses AI to answer customers but nobody has shown them where the line sits, that is exactly what my AI training for teams is built for. If you would rather find out where your business stands first, take the free Business AI Readiness Scorecard: 12 questions across six areas, about five minutes.

Sources and further reading

Dan Fitzpatrick helps businesses use AI well through keynotes, practical AI training for teams, and AI strategy and governance for leadership teams.

Key takeaways

  • AI helps most with the reading and writing in customer service, such as sorting, summarizing and drafting, and least with deciding.
  • In a Gartner survey published August 4, 2026, 87% of customers said companies using generative AI for service must offer a way to reach a human agent.
  • The same survey found customers were about three times more likely to use their own AI tools than a company's chatbot, so your written policies are now read by machines too.
  • In Moffatt v. Air Canada (2024), a Canadian tribunal held the airline responsible for its chatbot's wrong answer and rejected the idea that the chatbot was a separate legal entity.
  • The Promise Rule: AI may tell a customer anything the business has already written down, but may never promise anything a person has not already decided.
  • Start AI in the inbox, where a person reads every draft, before putting a chatbot on the website.
  • Keep AI away from customers entirely until your answers are written down, and in regulated advice.

Frequently Asked Questions

Should a small business use an AI chatbot for customer service?

Only for questions your business has already answered in writing, such as opening hours or order status, and only with a visible route to a person. Most small businesses get more value first from AI drafting inbox replies that a person checks, because every mistake is caught before a customer sees it.

Do customers mind talking to AI in customer service?

Many do not, as long as a person is reachable. A Gartner survey published in August 2026 found 50% of customers say interactions are easier when companies use generative AI, but 87% said a company using it must provide a way to reach a human agent.

Is my business liable if its AI chatbot gives a customer wrong information?

You should assume so. In Moffatt v. Air Canada in 2024, a Canadian tribunal held the airline responsible for its chatbot's wrong answer and rejected the argument that the chatbot was a separate legal entity. It is not binding in the UK or US, but the reasoning is common sense.

What customer service tasks can AI do in a small business?

AI can sort incoming messages, summarize long email threads, draft replies from your written policies, tidy phone notes into order updates and improve help pages. These are reading and writing jobs. Decisions such as refunds outside policy, delivery promises and complaints should stay with a person.

What is the Promise Rule for AI in customer service?

The Promise Rule is Dan Fitzpatrick's suggested line for small businesses: AI may tell a customer anything the business has already written down, but may never promise anything a person has not decided. Written answers can go to AI, promises are drafted by AI and sent by a person, and upset customers go to a person.

How should a small business start using AI for customer service?

Pull your last 50 customer messages and sort them by whether a written answer exists and whether the reply makes a promise. Write down your ten most common answers, then let the person who runs the inbox use AI to draft replies for two weeks, sending nothing unedited.

Dan Fitzpatrick helps businesses use AI well through keynotes, practical AI training for teams, and AI strategy and governance for leadership teams.

Dan Fitzpatrick

Written by

Dan Fitzpatrick

Dan Fitzpatrick is The AI Educator: a Forbes contributor, international keynote speaker and author who has trained more than 150,000 people across 30+ countries. He works with businesses and leadership teams on AI training, AI strategy and governance, and keynotes.

He also works with schools as an AI keynote speaker for schools, through school AI strategy consulting and with AI training for schools.