YES

As of 13 August 2026, AI can answer your customers' FAQs.

This still needs a person who signs their name to it.

Can you do it?

5 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ita colleague

What the alternative costsA purpose-built alternative is CustomGPT, which creates a no-code chatbot from your business content with citations.

If this goes wrong: customers receive an incorrect or overconfident answer, and your business has to repair the misunderstanding.

What to actually do

  1. Hand it to a person

    The route this page recommends

    A person who owns the outcome does this end to end, worth it when the failure is dear.

  2. Use a tool built for this

    Second choice

  3. Do it yourself

    The distant third

    A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Open your current FAQ, product pages, price list, delivery policy, refund policy, opening hours and escalation guidance, then remove superseded information.
    2. Paste those documents into the prompt under 'Business information' and add the business name and the customer support tone you want.
    3. Ask the chatbot to produce the table, including a source section and an escalation flag for every answer.
    4. Compare each drafted answer with the exact current source, especially prices, deadlines, eligibility rules, guarantees and refund wording.
    5. Rewrite any answer marked '[NEEDS HUMAN ANSWER]' or '[ESCALATE]' with the responsible person or team and the approved response.
    6. Give a colleague a sample of ordinary, ambiguous and difficult customer questions to test before publishing the approved answers or connecting them to live chat.

    Prompt

    You are drafting tier-1 customer-service FAQ answers for [BUSINESS NAME]. Use only the business information pasted below. Do not invent facts, prices, guarantees, delivery times, legal rights, product capabilities or exceptions. If the source does not answer a question, write '[NEEDS HUMAN ANSWER]' rather than guessing. Keep each answer between 40 and 90 words, use plain British English, answer the question first, and avoid marketing claims. Separate information that is certain from information that needs confirmation. Add '[ESCALATE]' to questions involving complaints, refunds outside the stated policy, safety, personal data, account access, legal threats, vulnerable customers or anything the sources do not clearly cover. Return a table with these columns: customer question, draft answer, source section, escalation needed. After the table, list every answer containing a price, deadline, eligibility rule or promise so I can check it separately.

Business information:
[PASTE CURRENT FAQs, PRODUCT DETAILS, POLICIES, PRICE LIST, DELIVERY INFORMATION, OPENING HOURS, CONTACT ROUTES AND ESCALATION RULES HERE]

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

What it gets wrong

Even on a YES, the friction has a name: context depth, verification cost and stakes of error.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs2
Verification1
Liability1
Effort delta2
Total8 / 10

The methodology and its thresholds are published in full.

FAQ

Can AI answer customer questions on my website?
Yes, AI can answer routine questions when it has accurate business content to use. Keep a human route for complaints, exceptions, sensitive account matters and anything the supplied information does not clearly answer.
How do I stop AI making up answers to customers?
Give it current source documents and instruct it not to guess, invent facts or fill gaps. Require a clear escalation label for unanswered questions, then compare every price, deadline, promise and eligibility rule with your approved information.
Can AI answer customer FAQs accurately?
It can handle straightforward FAQs accurately when the underlying content is complete and current. Accuracy falls when policies conflict, information has changed or the customer needs an exception, so approve the answers and test difficult examples before they go live.
What customer questions should AI not answer?
Do not let it own safety concerns, legal threats, sensitive complaints, personal-data issues, account-access problems or requests outside your stated policy. Route these to a trained person who can check the case and take responsibility for the response.

Nearby answers

All Tier-1 support & chatbots rulings

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.