How to remove your data from ChatGPT (OpenAI) in 2026

Updated Corrections

At a glance

  • Platform ChatGPT
  • Company OpenAI
  • Deletion and retention Deleted saved chats leave your account view immediately and are scheduled for system deletion within 30 days, with exceptions for already de-identified content and security or legal retention. Library files need separate deletion.
  • Official page OpenAI data controls

OpenAI documentation checked October 1, 2026 for the controls described here. Signed-in interfaces, request submission and removal outcomes were not tested. Availability depends on account, region and workspace.

What you need to know first

Training preferences, saved chats, memory, files and personal data in responses need separate actions. Choose the action that fits your concern; you do not need to delete an account to turn off model improvement.

Step-by-step: manage your ChatGPT data

  1. Step 1: Sign in at chatgpt.com and open your account menu > Settings.
  2. Step 2: For new eligible conversations, open Data controls and turn off Improve the model for everyone. This keeps your saved chats.
  3. Step 3: To remove a saved conversation, delete that chat. Archiving only moves it out of the sidebar.
  4. Step 4: Review memory under Settings > Personalization. Delete relevant memories and source chats separately; deleting a chat alone does not delete a saved memory.
  5. Step 5: Review saved files in Library and delete them separately. Project and custom GPT files follow their own retention rules.
  6. Step 6: For personal data appearing in a response, use OpenAI’s Privacy Portal and choose the personal-data response request. Keep the response, any cited URL and the reason for your request.
  7. Step 7: If you choose permanent account deletion, first review access and subscriptions. On the web, use Settings > Account > Delete account; if you cannot sign in, use the verified account-deletion request in the Privacy Portal. Managed accounts require your organization’s help.

On mobile

  1. In the iOS or Android app, open the sidebar and select your profile icon to open Settings.
  2. Open Data controls and turn off Improve the model for everyone. When signed in, this preference applies across devices.
  3. These mobile steps change training permission. Follow OpenAI’s device-specific account-deletion instructions if you want to delete the account.

Opt out of AI training

For a personal account, turn off Improve the model for everyone in Settings > Data controls, or select Do not train on my content in OpenAI’s Privacy Portal. Either route covers new ChatGPT conversations and personal-plan Codex tasks. Codex’s Include environments control is separate. Providing thumbs-up or thumbs-down feedback may make the associated conversation eligible for training even after opt-out.

Retention and deletion limits

Regular and archived chats remain until deleted or removed under a workspace policy. Files stored in Library are managed separately from chats. Files in a project or custom GPT remain until that project or GPT is deleted, then follow the documented deletion window and exceptions. Unsaved temporary chats may be retained for up to 30 days for safety; saving one makes it subject to regular-chat rules.

Separate data request

At privacy.openai.com, choose Make a Privacy Request and the option for removing personal data from ChatGPT responses. You can request this with or without a ChatGPT account. Provide specific examples, cited URLs and your reason; OpenAI requires identity verification and may ask a representative to prove authority. Use the official channel for verification documents. Requests are assessed individually and may be declined.

Key gotchas

What this doesn’t cover

A response can reflect learned patterns or a current web source. A personal-data response request does not remove external websites or search results. Seek removal at any cited source separately; an answer about you is not proof that your data was used in training.

Delist finds exposed personal data, pursues supported removal requests and tracks their status. Provider account controls and response requests may require your account access or identity verification. Removing a public listing does not establish that a trained model has changed.

See what personal data is exposed about you beyond ChatGPT.

Run a free scan →

Frequently asked questions

Does turning off training delete my ChatGPT chats?

No. The model-improvement preference covers new eligible content. Delete saved chats, files and memories through their separate controls. Feedback can make the associated conversation eligible for training even after opt-out.

Is a temporary chat always excluded from model training?

It is excluded while it remains temporary. If you save it, it becomes a regular chat and follows your account’s model-improvement and personalization settings.

Can a temporary chat use my existing memories?

Yes, if personalized. Choose Unpersonalized before the first message to turn off memory, custom instructions and plugins for that chat. Neither option creates or updates memories while temporary. Account restrictions and limited safety context still apply.

Does deleting a chat also delete its files and memories?

Not necessarily. Saved memories and Library files are separate from chat history. Review and delete them separately. Project and custom GPT files have their own retention rules.

How long can deleted ChatGPT content be retained?

Deleted saved chats leave your account view immediately and are scheduled for system deletion within 30 days, with exceptions for already de-identified content and security or legal retention. Library files need separate deletion.

Will a personal-data response request remove my data everywhere?

No. OpenAI reviews requests individually and may decline them. The request does not remove external websites or search results. Keep the specific response and cited source, and address the source separately.

Steps and URLs reflect the platform documentation as of the date above. Platforms may change their settings layout at any time. This guide is informational and not affiliated with OpenAI.

Check your public exposure

Delist finds exposed personal data, pursues supported removal requests and tracks their status. Provider account controls and response requests may require your account access or identity verification. Removing a public listing does not establish that a trained model has changed.

Run a free scan →