As of 13 August 2026, AI can collect contact details from customer enquiries.
Most people should hand this to a purpose-built tool.
Can you do it?
5 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0/month
Skill neededpower-user
Who has to check ityou
What the alternative costsThe supplied tool data lists Landbot, Botpress and Voiceflow as purpose-built platforms for building customer-facing AI chatbots, but gives no prices for them.
If this goes wrong, you may lose a genuine enquiry, store inaccurate personal data or create a privacy problem that your organisation must resolve.
What to actually do
Use a tool built for this
The route this page recommends
Hand it to a person
Second choiceA person who owns the outcome does this end to end, worth it when the failure is dear.
Do it yourself
The distant thirdA chat interface, power-user skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open your chatbot builder or customer-service platform and create a new enquiry flow with a handoff route to your staff inbox or CRM.
- Gather your business name, the contact fields you genuinely need, your response time, your human escalation route and the wording of your privacy notice.
- Paste the prompt into the chatbot configuration, replacing the bracketed slots with your approved business details and policies.
- Create saved fields for the visitor's name, email address, telephone number, enquiry summary, consent status, consent wording or version, and the time of consent.
- Connect the completed flow to the staff inbox or CRM, then submit a fictional enquiry and confirm that the fields, transcript and consent record arrive together.
- Run the five test conversations from the generated test set, correct any messages that request unnecessary information, and compare the consent wording with your current privacy notice before publishing.
- Publish the flow on the enquiry channel, tell staff how to handle handoffs, and set a retention and deletion process for the captured contact records.
Prompt
Design a customer-enquiry chatbot flow for [BUSINESS NAME]. Its main job is to collect contact details from people who want a reply. Ask only for the information needed to follow up: name, email address or telephone number, and a short description of the enquiry. Explain why each detail is needed before asking for it. Do not request passwords, payment details, health information or other sensitive information. Tell the visitor that a person may contact them, state the expected response time as [RESPONSE TIME], and ask for clear permission before storing or using their details for this enquiry. Do not treat that permission as marketing consent. Confirm each captured field back to the visitor, allow corrections, and end with a clear summary. If the visitor reports an urgent, unsafe or legally serious matter, stop collecting routine details and route the conversation to [HUMAN TEAM OR CONTACT ROUTE]. Produce: 1. the exact visitor-facing messages in order, 2. the fields to save, 3. the consent record to save, 4. the handoff message, 5. a JSON example of a completed record using fictional data, and 6. five test conversations covering a normal enquiry, an invalid email address, a correction, a refusal to provide details and an urgent request. Do not invent business policies, response times or contact routes.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot decide which personal data your organisation is legally justified in collecting or how long it should be retained.
- AI cannot guarantee that a visitor has entered a genuine or reachable email address or telephone number.
- AI cannot reliably recognise every urgent, vulnerable or unusual enquiry that needs a person immediately.
- AI cannot take responsibility for a data breach, an unlawful marketing message or a missed customer response.
- AI cannot replace the CRM connection, access controls and retention process needed after the conversation.
Even on a YES, the friction has a name: consent and privacy, context depth and real time truth.
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.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 2 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 9 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI collect customer names and email addresses?
- Yes. A configured chatbot can ask for names and email addresses, check basic formatting, confirm the entries and send them to your staff or CRM. You still decide why the data is collected and how it is stored.
- Can an AI chatbot collect contact details from a website enquiry?
- Yes. Website chatbot platforms can collect contact fields as part of an enquiry flow and hand the conversation to a person. Test the full path from the chat window to the staff inbox before publishing it.
- Is it legal to collect customer contact details with AI in the UK?
- AI does not make the collection lawful by itself. Give people a clear reason for collecting their details, collect only what you need, explain how it will be used and follow your UK data protection and retention processes.
- Can AI check whether a customer phone number is real?
- It can check whether a number follows a plausible format, but it cannot prove that the number belongs to the customer or that it can receive calls. A confirmation message or staff follow-up is needed when accuracy matters.
Nearby answers
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.