How to use and configure ChatGPT to protect your privacy

As artificial intelligence tools become more integrated into daily life, protecting your personal information when using ChatGPT is more important than ever. While complete privacy is difficult to guarantee, OpenAI offers a number of options that allow you to limit data sharing and maintain better control over your information.

Use temporary chats for sensitive conversations

Temporary chats do not save your conversation history or contribute to model training, making them ideal for private or sensitive topics. This feature works much like an incognito window in a browser.

Disable data sharing for model improvement

By default, ChatGPT uses your conversations to improve its models. However, in your account settings under Data controls, you can disable the “Improve model for everyone” option to stop OpenAI from using your input for training.

Turn off memory-based personalization

ChatGPT can remember details from past chats to personalize your experience. To avoid references to older conversations, go to Personalization and turn off the memory feature.

Manage or delete saved memories

Under Personalization settings, you can access and delete specific memories ChatGPT has saved, or clear all stored data to maintain better privacy.

Delete or archive chat history

If you’re concerned about others seeing past conversations, you can delete or archive any chat. Archiving hides it from view without deleting it entirely.

Monitor shared links and uploaded files

Be cautious with shared conversations. You can manage all shared links under Data controls. Also, avoid uploading files with sensitive metadata or personal content.

Avoid unnecessary integrations

Limit third-party app and service integrations, especially those involving work data or cloud storage, to reduce the risk of external data access.