Is ChatGPT private? What OpenAI keeps and how to opt out
ChatGPT is private from other users unless you share a chat, but it is not private from OpenAI. On personal Free, Plus and Pro accounts, OpenAI can use your conversations to train its models by default, and it keeps your chats until you delete them.
That doesn't make ChatGPT unsafe to use. It means a few settings matter, and a few habits matter more. Here is what OpenAI's own privacy policy and help pages say, as of November 2026, and the switches worth changing.
What ChatGPT collects
OpenAI's privacy policy, updated September 10, 2026, lists what it gathers when you use ChatGPT:
- Account details: "your name, contact information, account credentials, date of birth, payment information, and transaction history."
- Your content: prompts and anything you upload, "such as files, images, audio and video, and data from connected services."
- Log data: your IP address, browser type and settings, the time of each request and "how you interact with our Services."
- Location: the general area you connect from, based on your IP address, and more precise GPS location only if you choose to share it.
- Contacts: only if you connect your device's address book, in which case OpenAI checks which of your contacts also use its services.
- Ad data, on Free and Go: "information about your ads history and interests," plus information from advertisers, such as purchases you make with them.
The policy also says OpenAI may monitor content "to prevent fraud, illegal activity, or misuses of our Services," and that it does not "sell" personal data.
Does ChatGPT train on your data?
Yes, by default, on personal accounts. OpenAI's data controls page says: "On personal workspaces with ChatGPT Free, Plus, or Pro, data sharing is enabled by default." That covers your chats and, according to the Memory FAQ, the things ChatGPT remembers about you.
OpenAI says it takes "steps to reduce the amount of personal information in training datasets" before training. Business plans work differently: by default, OpenAI does not train on content from ChatGPT Business, Enterprise, Edu or ChatGPT for Healthcare, or on the API.
How to turn off "Improve the model for everyone"
This is the main ChatGPT privacy setting. Turning it off keeps your chat history; it only stops new conversations being used for training.
On the web
- Open your account menu.
- Select Settings.
- Select Data controls.
- Select Improve the model for everyone, turn it off, and select Done.
On iPhone and Android
- Open the sidebar.
- Select your profile icon to open Settings.
- Select Data controls.
- Turn off Improve the model for everyone.
When you're signed in, the choice applies to your account on web and mobile. If you use ChatGPT signed out, open Settings and turn it off there, but it is only saved in that browser.
You can also select Do not train on my content in OpenAI's Privacy Portal at privacy.openai.com. OpenAI's model training page says "either option is sufficient," and after you opt out, "we won't use your new conversations to improve our models."
Three notes: Voice mode has its own separate training control, so check it if you talk to ChatGPT. Archiving a chat does not change whether it can be used for training. And OpenAI says that during some special Trusted Access sessions on a personal account, it may use content to improve its models even with the setting off; the setting still covers the rest of your activity.
The feedback exception
This one catches people out. Even with training off, OpenAI says: "If you choose to provide feedback, such as selecting thumbs up or thumbs down on a response, the entire conversation associated with that feedback may be used to train OpenAI models."
So a quick thumbs up on a chat about your health or your finances shares the whole conversation, not just that one reply. Skip the rating buttons on anything personal. Conversations with OpenAI Support may also be used to improve its services if the training switch is on.
Temporary Chat: what it does and doesn't do
Temporary Chat is ChatGPT's closest thing to a private mode. According to the Temporary Chat FAQ, temporary chats don't appear in your history, don't create or update memories, and are not used to train OpenAI's models.
To start one, open a new chat, select Temporary, then choose Personalized or Unpersonalized before your first message. Unpersonalized ignores your memories and custom instructions; Personalized can use them. You can't switch after the chat starts.
The catch: "OpenAI may keep a copy of a temporary chat for up to 30 days for safety purposes." And if a custom GPT sends data to an outside service, that service's own privacy policy applies, and it may keep the data longer.
Memory and connected apps
Memory is how ChatGPT personalizes answers, and OpenAI says it is a separate setting from training. Memory can draw on past chats, saved memories, custom instructions, saved files and "content from connected apps, such as Gmail."
To manage it, open Settings, select Personalization, then Memory. A few details from the Memory FAQ are worth knowing:
- Turning off Memory does not delete your past chats.
- "Deleting a chat alone does not necessarily delete a separate saved memory created from that chat." Delete the memory too.
- If you turn off Reference chat history, what ChatGPT remembered from past chats is scheduled for deletion within 30 days.
- "Disconnecting an app prevents future access to that app. It does not delete past conversations that already used app content."
If you've connected your email, ChatGPT can read what's in it when it answers. We cover that wider question in how to stop AI from reading your emails.
Deleting your chats and your account
Signed-in Free, Go, Plus and Pro users can export their data and delete their account from ChatGPT settings. Deleting your OpenAI account is permanent.
The privacy policy says that once you delete data, OpenAI will "remove it from our systems within 30 days," with two exceptions: data it needs to keep longer, for example under "a lawful subpoena," and content that "has already been de-identified and disassociated from your account" because you allowed training. In plain terms, turning training off early protects you more than deleting later.
Ads on Free and Go
On the Free and Go plans, OpenAI uses your ads history and interests "to personalize the ads you see on our Services (subject to your settings)" and to measure ads. You can opt out of targeted advertising with the marketing privacy control in your account settings. If you're not logged in, the policy points to Settings > Data Controls in ChatGPT. OpenAI says it doesn't do this for users it knows are under 18.
Is ChatGPT private at work?
From OpenAI, it's more private by default: Business, Enterprise and Edu content isn't used for training unless your company changes that. From your employer, no. The privacy policy says that when you join a ChatGPT Enterprise or business account, "the administrators of that account may access and control your OpenAI account, including being able to access your Content."
Treat a work ChatGPT like work email. If you're curious how far that goes, see can my employer read my email?
Frequently asked questions
Is ChatGPT private and secure?
Other users can't see your chats unless you share a link. OpenAI can, and its policy says it uses "commercially reasonable" security measures but that "no Internet or email transmission is ever fully secure," so it advises "special care in deciding what information you provide."
Does ChatGPT share your conversations with the government?
The privacy policy says OpenAI may share personal data, including your interactions, with government authorities "in compliance with the law," for example "if required to do so to comply with a legal obligation." It may also keep data for as long as a legal obligation such as a subpoena requires.
Does ChatGPT save my data if I turn off training?
Yes. Turning off Improve the model for everyone does not delete or hide your saved chats. Delete them yourself, or use Temporary Chat for things you don't want kept.
Is ChatGPT safe for personal information?
Use the settings above, and still keep passwords, card numbers, ID numbers and other people's private details out of the chat box. Even temporary chats can be kept for up to 30 days for safety.
How does ChatGPT compare with other AI assistants?
Each one handles training and history differently. See our guides to Claude, Gemini and Copilot, or the round-up on how to opt out of AI training.
Keep your inbox out of the ad business
A lot of what people paste into ChatGPT starts in their email: receipts, bank letters, messages from family. If you want that inbox to be more private too, Secria Mail has no ads and Secria never mines your mail. Your mailbox is locked with a key made from your password, on your device, and sign-up is a name, the address you want and a password, with no phone number and no card.
Secria fact-checks every post against primary sources. Spotted something wrong or out of date? Email hq@secria.me and we will correct it.