---
Title: "What Is Intent Detection? Definition and Examples"
Url: "https://computer.io/glossary/intent-detection"
Topic: "AI foundations"
Description: "Intent detection is AI that identifies what a user wants from their message, like a refund or a password reset. How it works and how it's used in support."
---

# What is intent detection?

Intent detection is AI that identifies what a person wants to accomplish from their message, such as requesting a refund, resetting a password or reporting a bug, so a system can route, answer or act on it correctly.

## Example

| Message | Intent |
| --- | --- |
| "I was charged twice" | Billing dispute |
| "Can't log in" | Account access |
| "How do I export?" | How-to |
| "Cancel my plan" | Cancellation |

## Old vs new approaches

Classic chatbots needed predefined intents and training phrases. LLMs can infer intent from open-ended language without a fixed list, but defined categories still help reporting.

## Uses

Routing, triage, analytics and triggering automations.

## FAQ

### What is intent classification?

Another name for intent detection: assigning a message to an intent category.

### Do LLMs need intent training?

Not always; they can classify with instructions and examples.

### What is multi-intent detection?

Recognizing more than one goal in a single message.

## Related terms

- [Ticket triage](https://computer.io/glossary/ticket-triage)
- [Sentiment analysis](https://computer.io/glossary/sentiment-analysis)
- [Conversational AI](https://computer.io/glossary/conversational-ai)

## How Computer helps

- [Customer agent](https://devrev.ai/solutions/customer-agent)

## Further reading

- [Ai support ticket triaging](https://devrev.ai/blog/ai-support-ticket-triaging)