Jump to a section
ChatGPT is an AI chatbot made by OpenAI. You type or say what you want, and it writes a reply: explaining things, drafting messages, planning, brainstorming. It's brilliant with words but can be confidently wrong, so check anything important. The free plan is enough for most people to start, and a couple of settings keep your chats more private.
ChatGPT in one sentence
ChatGPT is an AI chatbot made by the company OpenAI: you write to it in ordinary language, and it writes back.
Behind it is what's called a large language model. OpenAI explains that it learns patterns from huge amounts of text, then answers by predicting 'the most likely next piece of language based on context'. It doesn't look facts up in a book, and OpenAI is clear it 'doesn't "know" things the way a person does'.
How AI chatbots actually work, and why it changes how you use them
What it's good for
Think of it as a patient, well-read helper for anything involving words and ideas. People use it to:
- Explain something confusing, like a letter, a bill or a news story
- Write or tidy up an email, a complaint or a card message
- Plan meals, trips, parties and budgets
- Brainstorm gift ideas, names or things to do
- Practise a language, or get homework explained (not done)
What it gets wrong
Because it predicts likely words, it can sound completely sure about something false. Researchers at OpenAI have argued that the way these systems are tested rewards guessing over admitting uncertainty. The more an answer matters, the more you should check it.
AI chatbots make things up. Here's how to check what they tell you
OpenAI: Why language models hallucinate (5 September 2025) Kalai and colleagues: Why Language Models Hallucinate (research paper, 4 September 2025)
Does it know what happened today?
Not always. A model only learns from text up to a certain date, and while ChatGPT can search the web, it doesn't always choose to. The Columbia Journalism Review described ChatGPT dismissing a major news story hours after it happened, then answering correctly days later when it searched. For anything recent, ask it to search and show its sources.
Is it free?
Yes, there's a free plan, as well as paid ones. Ads appear on the Free plan and the Go plan; OpenAI says Plus and Pro don't have them. Ads were switched on in the UK in June 2026, as PPC Land reported. OpenAI says advertisers don't get access to your chats.
Most people can start, and stay, on the free plan. Paying makes sense when you're relying on it for bigger jobs.
Two privacy controls worth knowing
For ordinary personal chats, open Settings, then Data controls, and review Improve the model for everyone. Turn it off if you do not want new conversations used for model improvement. That does not delete your saved history. If you send feedback on a response, the associated conversation may still be used for training.
For a conversation you do not want saved in history, start a new chat and select Temporary. Choose its personalisation option before sending your first message. It is not used for model improvement while it stays temporary, but OpenAI may retain a copy for up to 30 days for safety.
If you later save that temporary conversation, it becomes a regular chat and follows your normal settings. These controls do not make a personal chatbot a confidential advice service, so keep private details out.
Your first go
Go to chatgpt.com or download the official ChatGPT app from your phone's app store, and sign up. Then try something low-stakes and useful, like this:
I've got chicken, peppers, rice and a tin of tomatoes. Suggest two easy dinners I could make tonight in under 30 minutes, with short steps.How to ask an AI chatbot a good question (the four-part trick)
How to make ChatGPT, Gemini and Claude answer for your country
Can you still use ChatGPT on WhatsApp? No, here's what to use instead
Sources
These sources support the guide. Tool settings, prices and policies change, so dated source boxes show when those details were checked.
- OpenAI: testing ads in ChatGPT (updated 7 May 2026)
- PPC Land: ChatGPT ads go live in the UK as OpenAI expands pilot beyond US (June 2026)
- Columbia Journalism Review, Tow Center: AI chatbots can search the web, so why don't they? (27 January 2026)
- OpenAI Academy: AI fundamentals (10 April 2026)
- IBM: what is ChatGPT? (19 April 2025)
- Giles Thomas: what AI chatbots are actually doing under the hood (29 August 2025)
- TDWI: context windows in AI (May 2026)
- OpenAI: Why language models hallucinate (5 September 2025)
- Kalai and colleagues: Why Language Models Hallucinate (research paper, 4 September 2025)
- The Register: OpenAI says models are programmed to make stuff up instead of admitting ignorance (17 September 2025)
- OpenAI: Data controls in ChatGPT
- OpenAI: Temporary chat in ChatGPT
AI helps with research and drafting. Dan checks the sources and is responsible for what is published.
How we check