How it works
Users type prompts (ideas or questions) and the app replies. Its responses may not be factually correct. Users can also speak to the app, but ‘Advanced Voice Mode’ provides a more realistic, real-time conversation where the app reacts to tone and emotion.
There are no adverts. The app can be personalised, such as using more friendly and enthusiastic language vs a calmer and more neutral tone. If the Memory feature is switched on, the AI remembers details about the user, like names or preferences, to make future chats more helpful. Chat history is not saved if using Temporary Chat Mode.
A user can use a GPT. These are custom, mini-versions of the AI built for one specific purpose, like a personal tutor or a cooking assistant. A user can invite up to 20 people to join a group chat conversation with the AI. AI models use a lot of energy, so you can reduce their environmental impact by keeping prompts short and creating images only when necessary.
Parental controls or safety settings available
Users can report harmful content. Teenage accounts have some default safety settings e.g. sensitive content filters.
Optional parental controls allow parents to link their account to their child’s (13-17) with the teenager’s agreement. This gives parents the ability to:
• Set usage times
• Disable voice chat, image generation, memories and group chat
• Adjust the content filters
• Stop their child’s data being used to train the AI
Parents cannot see their child’s chat history, but will be alerted if the app detects acute distress (e.g. talking about self-harm). Both the app and web versions of ChatGPT can be used without logging in. There are filters applied automatically to reduce the likelihood of harmful content.
