Chatbot

A chatbot is a piece of software that holds a text-based conversation with a user, responding to questions and requests in plain language. Early chatbots worked by matching keywords to pre-written answers, which made them brittle and easy to confuse. Modern AI chatbots are powered by large language models, which means they can understand context, handle follow-up questions, and generate flexible responses to almost anything a user types.

How it works#

When you send a message to an AI chatbot, your text is broken into tokens and processed by a language model that has been trained on a vast amount of text. The model predicts the most appropriate next sequence of words based on your input and the conversation history so far. This is not a simple lookup from a database. The model is generating a response on the fly, which is why modern chatbots feel much more natural than the older rule-based systems. The underlying process is explained in more detail in how LLMs work.

Most chatbot products also layer additional instructions on top of the base model. These system-level instructions tell the chatbot to stay on topic, maintain a certain tone, avoid certain subjects, or follow a particular format. This is done through prompt engineering.

Examples#

Customer service on a retail website. When you visit a clothing retailer and click the chat icon, a chatbot can instantly tell you the status of your order, explain the returns policy, or help you find the right size. It handles the most common queries without needing a human agent, which means you get a faster answer at any time of day.

AI writing assistants. Products such as ChatGPT and Claude act as conversational tools for drafting emails, summarising documents, or brainstorming ideas. You type what you need the way you would say it, the chatbot produces a draft, and you can refine it by replying with follow-up instructions. Our beginners guide to Claude walks you through this kind of interaction step by step.

Why it matters#

Chatbots have moved well beyond novelty. They are now handling millions of interactions per day across healthcare, banking, education, and retail. For individuals, they offer an always-available resource for getting quick answers or drafting text. For organisations deploying them, they can handle routine enquiries at scale, freeing up staff for work that genuinely requires human judgement.

Understanding how chatbots work also helps you use them more effectively. Knowing that they rely on natural language processing and pattern matching -- rather than genuine understanding -- explains why clear, specific prompts produce better results than vague ones.

To see chatbots in practice, start with our ChatGPT beginners guide or browse the full AI tools collection.