AI Privacy Basics for Beginners
Assume anything you type into a chatbot could be stored, reviewed or used to improve the product unless you’ve checked the settings. Don’t paste passwords, bank details, or other people’s private information.
Never paste these
- Passwords, security codes and recovery phrases.
- Full card or bank details.
- National Insurance / social security numbers, passport or ID details.
- Other people’s private information (customers, patients, colleagues) without permission.
- Confidential work documents your employer hasn’t approved for AI use.
Check your settings
Each provider lets you control some things, such as whether chats are used to improve their models, whether chat history is saved, and whether memory is on. The options and their names differ by provider and plan and change over time, so look in the settings of ChatGPT, Claude or Grok and read the provider’s own privacy page.
Simple habits
- Anonymise. Swap real names for “Client A” before pasting.
- Share the minimum. Paste the paragraph you need, not the whole file.
- Use work-approved tools for work data.
- Use a strong, unique password and two-factor authentication on your AI accounts.
- Be careful with agents and browser extensions that ask for account access. See What are AI agents?
Be wary of sites promising “free ChatGPT Plus” or “unlimited Claude Pro”. They’re often scams that harvest logins. Use the official sites only.
What to try next
Back to basics with What is AI?
Sources
- General guidance written by Silent Overseer. Each provider’s privacy settings differ and change: check ChatGPT, Claude and Grok’s own privacy pages.
- ChatGPT
- Claude
- Grok
Silent Overseer uses AI tools to help research and draft, and every article is reviewed before publishing. How we work. Spotted a mistake? We correct and date changes.