The Beginner's Guide to Claude

If you have heard people talk about "Claude" and felt a step behind, this guide is for you. There is nothing to be nervous about. Claude is a computer program you talk to in plain language — you type a question or a request, and it writes back. That's it. By the end of this page you'll know what it is, what it's good and bad at, and exactly how to try it well.

What Claude actually is

Claude is an AI assistant made by a company called Anthropic. Under the bonnet it is a large language model — a program trained on an enormous amount of text so that, given some words, it can predict sensible words to follow. In practice that makes it very good at reading, writing, summarising, explaining and drafting.

Claude is a family of AI models built by Anthropic, available to chat with on the web and in apps.

You can use it at claude.ai in a web browser, or through Anthropic's apps. There is a free tier you can start on today, and paid plans that give you more usage and access to the more capable models. You do not need a paid plan to learn what it does.

A helpful mental model: Claude is a very well-read, very fast assistant who is sometimes confidently wrong. Brilliant for a first draft or an explanation. Never to be trusted blindly on a fact that matters.

What Claude is genuinely good at

These are the tasks where Claude earns its keep for an ordinary person or a small business:

• Explaining things simply. Paste a confusing letter, contract clause or error message and ask, "What does this mean in plain English?"
• Drafting writing. Emails, adverts, job descriptions, a polite complaint, a speech for a retirement do. You give the gist; it gives you a first draft to edit.
• Summarising. Long documents, reports, or a wall of meeting notes into the three things that matter.
• Rewriting and tidying. Making something shorter, warmer, more formal, or fit a word count.
• Thinking things through. "Here's my situation — what am I not considering?" It's a good sounding board, not a decision-maker.
• Working with your own documents. You can attach files (like a PDF or an image) and ask questions about them.

What Claude is bad at — the honest limits

This is the part the hype skips. Knowing the limits is what keeps you safe and saves you embarrassment.

• It can make things up. When it doesn't know, it may produce a fluent, confident answer that is simply wrong — a made-up statistic, a fake quote, a book that doesn't exist. This is often called a "hallucination." Treat every factual claim as unverified until you've checked it.
• It isn't a live search engine by default. Its core knowledge was fixed at a training cut-off, so it may not know recent events. Some versions can browse the web, but don't assume it's up to date unless it shows you a source.
• It has no real-world judgement or accountability. It can't know your business, your family or your legal situation. It doesn't understand the way a person does — it predicts text.
• It is not a substitute for a professional. For medical, legal, financial or tax decisions, use it to prepare questions — not to get the answer.

How to start, in five minutes

1. Go to claude.ai and create a free account.
2. In the message box at the bottom, type a normal request — see the examples below.
3. Read the reply. If it's not quite right, just say so: "Make it shorter," or "That's too formal," or "You got the date wrong — it's 2019."
4. Keep going in the same conversation. Claude remembers what you said earlier in that chat, so you can refine rather than start over.

That back-and-forth is the whole skill. You are having a conversation, not casting a spell.

Talking to it well (prompting, plainly)

"Prompting" just means how you ask. You don't need clever tricks. A good request usually has three plain parts: who it's for, what you want, and any facts it needs.

Compare these two:

• Weak: "Write an email about the delivery."
• Strong: "Write a short, friendly email to a customer named Sara. Her sofa delivery is delayed by a week because of a supplier problem. Apologise, give the new date (14 July), and offer £30 off. Keep it under 120 words."

The second one gets a usable draft first time, because you gave it the facts and the shape. A few habits that help:

• Give context. Tell it who you are and who the reader is.
• Say the format. "A bullet list," "a table," "under 100 words," "in British English."
• Show an example of the tone or style you want, if you have one.
• Ask it to ask you. "What do you need to know from me to do this well?" is one of the most useful sentences you can type.

Privacy — what you should know

This matters, so here is the honest version. Anything you type into an AI assistant leaves your device and is processed on the company's computers. So the simple rule is: don't paste in anything you wouldn't email to a third party — no passwords, no full bank details, no other people's private information.

Anthropic publishes a privacy policy and consumer terms explaining what it does with your conversations, including your choices about whether they can be used to improve its models. These settings and policies change, so check the current ones rather than trust a rumour.

Anthropic documents how it handles your data, and consumer settings let you control some of it — read the current policy rather than assume.

For anything sensitive — a business's customer data, health information, legal matters — check the settings, consider a paid or business plan with stronger data terms, and when in doubt, leave it out.

A fair word on the alternatives

Claude is one of several good assistants. ChatGPT (from OpenAI) and Gemini (from Google) do broadly similar things, and which one "wins" changes month to month. We name them plainly and we earn nothing if you pick any of them — that's the point. Try one, get comfortable, and don't agonise: the skill of asking well carries across all of them.

The honest bottom line

Claude is a fast, capable, genuinely useful assistant for reading, writing and thinking — and a confident liar about facts you haven't checked. Use it as a tireless first-drafter and explainer, keep yourself firmly in the editor's chair, and never paste in what you wouldn't want a stranger to see. Do that, and it will save you real time without ever putting you at risk.