aiessential.aiStart AI Basics
Menu
← Safety Hub

AI and your privacy: the settings worth checking today

5 min read · aiessential.ai

"Is AI safe?" is one of the most common questions beginners ask. We have written separately about whether ChatGPT is safe for beginners. This guide focuses on one practical part of that question: your privacy. What happens to the things you type, and which settings are worth five minutes of your time?

What happens to what you type

When you use an AI chat tool, your message is sent over the internet to the company's computers, processed there, and the reply is sent back. That means:

  • the company stores your conversations, usually for a period set out in its privacy policy
  • depending on your settings, your conversations may be used to train and improve future AI models
  • in some cases, trained staff or contractors may review conversations, for example to improve quality or check for misuse

None of this is secret. It is in the companies' own privacy policies and help pages. But few people read them, so it is worth knowing the headlines.

Your rights in the UK

In the UK, the use of personal data is governed by the UK GDPR and the Data Protection Act 2018, overseen by the Information Commissioner's Office (ICO). Among other rights, you generally have the right to:

  • be told how your data is used
  • ask for a copy of the personal data a company holds about you
  • ask for data to be corrected or, in some circumstances, deleted
  • object to some kinds of processing

If you think an organisation has mishandled your data, you can complain to it first and then to the ICO. The ICO website (ico.org.uk) explains how.

The settings worth checking

Menus change from time to time, so the exact wording may differ slightly. These were correct according to each company's help pages at the time of writing.

ChatGPT (OpenAI)

  • Model training: go to Settings → Data controls and look for "Improve the model for everyone". Switch it off if you don't want your chats used for training.
  • Temporary chat: start a temporary chat for anything you would rather not keep. OpenAI says these are not saved to your history or used for training, though it may keep a copy for up to 30 days for safety reasons.
  • Memory: ChatGPT can remember details about you between chats. You can view, delete or turn off memory in Settings → Personalization → Memory (OpenAI's menus use US spelling).
  • Delete chats: you can delete single chats or clear all of them in settings.

Google Gemini

  • Activity: Gemini has an activity setting (now called "Keep Activity", and managed on the Gemini Apps Activity page) in your Google account. Google says that when it is on, your chats may be used to improve its services and some may be reviewed by human reviewers. You can turn it off and choose how long to keep activity.
  • Google's warning: Google's own privacy hub advises people not to enter confidential information they wouldn't want a reviewer to see.

Microsoft Copilot

  • Personalisation and training: in Copilot's settings, under privacy, you can control whether your conversations are used for model training and personalisation.
  • Work accounts: if you use Copilot through an employer's Microsoft 365 account, different rules apply, set by your organisation. Ask your IT team.

Voice assistants and smart speakers

Alexa, Google Assistant and Siri also use AI. Each has settings for whether voice recordings are kept and reviewed. It is worth looking in the app's privacy section.

The golden rule: what never to share

Settings help, but the strongest protection is simply not typing sensitive things in the first place. Keep these out of AI chats:

  • passwords, PINs and one-time security codes
  • full bank or card details
  • your National Insurance number, passport or driving licence number
  • detailed medical records, yours or anyone else's
  • other people's private information without their permission
  • confidential work documents, unless your employer has approved the tool

You can still get help with personal tasks. Use placeholders: "Help me write to [bank] about a payment of [amount] on [date]."

Photos and documents

Many AI apps let you upload pictures and files. Before you do:

  • cover or crop out names, addresses, account numbers and faces you don't need
  • remember that photos can contain hidden location data. Many phones let you remove this when sharing
  • think twice before uploading photos of children

Watch out for fake AI apps and sites

Some privacy risks come not from the real AI companies, but from imitators. Copycat apps and websites may use names like "ChatGPT" to harvest your data or charge for subscriptions.

  • Download apps only from official app stores, and check the publisher name (OpenAI, Google, Microsoft).
  • Be suspicious of any "AI tool" that asks for card details before you have even tried it, or asks you to log in with your bank.
  • The National Cyber Security Centre (NCSC) offers free, practical advice on staying secure online, including using strong passwords and two-step verification.

Protect the account itself

Your AI account may hold months of conversations, so protect it as you would your email:

  • use a strong, unique password (the NCSC recommends three random words)
  • turn on two-step verification (sometimes called multi-factor authentication) where available
  • sign out on shared computers

A five-minute privacy check

  1. Open your AI tool's settings and find the data or privacy section.
  2. Decide whether you want your chats used for training, and set it.
  3. Check whether memory or personalisation is on, and review what it has stored.
  4. Delete any old chats containing anything sensitive.
  5. Turn on two-step verification.

Using AI with confidence

Privacy concerns shouldn't put you off AI altogether. Used sensibly, these tools are genuinely helpful. It's about knowing where your information goes and making a few deliberate choices.

Our AI Basics course covers safe, confident everyday use, including privacy habits, spotting fake tools and avoiding scams, in short visual lessons with knowledge checks and a certificate of completion. Take a look at AI Basics.

Sources

  • ICO, Your data matters: https://ico.org.uk/for-the-public/
  • ICO, Guidance on AI and data protection: https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/
  • OpenAI, Data controls in ChatGPT: https://help.openai.com/en/articles/7730893-data-controls-in-chatgpt
  • OpenAI, Temporary Chat FAQ: https://help.openai.com/en/articles/8914046-temporary-chat-faq
  • OpenAI, Memory in ChatGPT: https://help.openai.com/en/articles/8590148-memory-in-chatgpt
  • Google, Gemini Apps Privacy Hub: https://support.google.com/gemini/answer/13594961
  • Microsoft, Copilot privacy controls: https://support.microsoft.com/en-us/topic/privacy-faq-for-microsoft-copilot-27b3a435-8dc9-4b55-9a4b-58eeb9647a7f
  • NCSC, Top tips for staying secure online: https://www.ncsc.gov.uk/collection/top-tips-for-staying-secure-online