Home · Business · Customer Service · Tier-1 support & chatbots
As of 13 August 2026, AI can hand a chat conversation to a human agent.
Most people should hand this to a purpose-built tool.
Can you do it?
30 minutesto a draft.
1 hourto something you’d act on.
Cost, all in£0/month
Skill neededpower-user
Who has to check ityou
What the alternative costsNo alternative price is provided in the supplied tools data.
If this goes wrong: the customer reaches nobody or has to repeat the issue, and your business carries the service and reputational cost.
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 1 hour until you can act on the result.
How to actually do it
- Open your chatbot or customer-support platform and confirm that it supports live-agent transfer, transcript sharing and an agent inbox or helpdesk connection.
- Gather your current escalation rules, support opening hours, agent queues, fallback contact route and the customer information your business is permitted to pass to an agent.
- Paste the prompt into the chatbot builder and replace the bracketed sections with your business details and escalation conditions.
- Configure each escalation condition from the returned workflow, including explicit requests for a person, complaints, repeated failed answers and sensitive account actions.
- Set the handoff message, waiting message, outside-hours message and no-agent fallback in the live-chat workflow.
- Run a test chat for every condition in the test plan, then open the receiving agent inbox and compare the transcript, customer details and escalation reason with the original chat.
- Send the workflow to a colleague to test from a customer account, fix any missing context or incorrect routing, and only then publish it to your website or messaging channel.
Prompt
Design a human-handoff workflow for a customer-support chatbot used by [BUSINESS NAME]. The chatbot should transfer the conversation to a human agent when [ESCALATION CONDITIONS], including requests for a complaint, sensitive account action, repeated failure to answer, or a customer explicitly asking for a person. Preserve the full transcript, the customer's contact details if already collected, the reason for escalation and any relevant conversation metadata. State what the customer should see while waiting, what happens outside agent hours, and how the chatbot should respond if no agent is available. Do not invent integrations, capabilities, response times or contact details. Return: 1) the workflow logic, 2) the information passed to the agent, 3) the customer-facing messages, 4) the failure and fallback paths, and 5) a test plan with sample conversations covering every escalation condition. Use plain British English.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- Cannot decide reliably whether an unusual or emotionally charged message needs a person unless you define the escalation rules and test edge cases.
- Cannot create agent availability, helpdesk permissions or channel integrations that your platform does not support.
- Cannot guarantee that the receiving team will respond, especially outside opening hours or when the queue is full.
- Cannot take responsibility for a customer being misrouted, left waiting or given an inappropriate automated message.
Even on a YES, the friction has a name: real time truth, stakes of error and context depth.
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 transfer a chatbot conversation to a human?
- Yes. A configured chatbot can recognise escalation rules, pass the transcript to an agent inbox and tell the customer that a person is taking over. You must test the transfer in the actual channel before publishing it.
- How do I make my chatbot hand over to a human?
- Connect the chatbot to a live-chat inbox or helpdesk, then define the conditions that trigger escalation and the messages shown while the customer waits. Test the transcript, customer details and escalation reason from the agent's side.
- Can a chatbot transfer a customer to a human outside business hours?
- It can route the conversation to a queue or collect a message, but it cannot make an unavailable agent respond. Set a clear outside-hours message, record the conversation and provide the approved fallback contact route.
- Will the human agent see the whole chatbot conversation?
- They can if the platform and integration pass the transcript during handoff. Check this with a real test conversation, because a transfer that opens a new ticket without the earlier context still makes the customer repeat the issue.
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.