Does ChatGPT (Free / Plus / Pro) train on your data?
Trains by default
Trains on business data unless you opt out.
Free, Plus and Pro: OpenAI may train on your chats unless you turn off "Improve the model for everyone". Opt-out covers new chats only. Thumbs up/down feedback can still send the whole conversation to training even after opting out.
Why this matters: Data used for training can't be pulled back out of a trained model, so opting out only protects what comes after. Why training matters →
Covers: Free, Plus, Pro. Compliance fields (SOC 2, ISO, DPA, BAA) have not been checked yet for personal plans.
Personal vs business plan
The same assistant can have different data terms depending on the plan. Here is ChatGPT (Free / Plus / Pro) next to ChatGPT Enterprise / Business.
| What we checked | Personal ChatGPT (Free / Plus / Pro) | Business ChatGPT Enterprise / Business |
|---|---|---|
| Trains on your data by default | Yes, opt-out | No |
| Data retention | Until you delete; deleted chats purged within 30 days | Admin-configurable |
| Data location / residency | US and other countries (no consumer region choice stated) | Yes |
| SOC 2 Type II | Not checked yet | Yes |
| DPA available | Not checked yet | Yes |
| BAA / HIPAA | Not checked yet | Yes |
See the business plan: ChatGPT Enterprise / Business → Full side-by-side comparison
Trains on your data by default
Free, Plus and Pro: OpenAI may train on your chats unless you turn off "Improve the model for everyone". Opt-out covers new chats only. Thumbs up/down feedback can still send the whole conversation to training even after opting out.
- help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance · checked Oct 17, 2026
“When you use our services for individuals (Free, Plus, Pro), we may use your content to train our models. You can choose whether your conversations help improve our models. To opt out, turn off Improve the model for everyone under Settings > Data controls in ChatGPT”
- help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance · checked Oct 17, 2026
“If you choose to provide feedback, the entire conversation associated with that feedback may be used to train our models.”
Can you turn training off?
Settings > Data controls > Improve the model for everyone (off), or 'Do not train on my content' in the OpenAI Privacy Portal. Applies to new conversations.
- help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance · checked Oct 17, 2026
“After you opt out, we won’t use your new conversations to improve our models.”
- help.openai.com/en/articles/7730893-data-controls-faq · checked Oct 17, 2026
“Select Improve the model for everyone, turn it off, and select Done.”
Data retention
Saved and archived chats stay until you delete them. Deleted chats are removed from view immediately and purged within 30 days, unless already de-identified or kept for security/legal reasons.
- help.openai.com/en/articles/8809935-deleting-and-archiving-chats-in-chatgpt · checked Oct 17, 2026
“Regular and archived chats remain saved to your account until you delete them or an applicable workspace retention policy removes them.”
- help.openai.com/en/articles/8809935-deleting-and-archiving-chats-in-chatgpt · checked Oct 17, 2026
“OpenAI schedules it for permanent deletion from its systems within 30 days, unless it was already de-identified and disassociated from your account or OpenAI must retain it longer for security or legal obligations.”
Retention after you delete
Content already de-identified for training is not removed by deleting the chat.
- openai.com/policies/us-privacy-policy/ · checked Oct 17, 2026
“Once you choose to delete Personal Data, we will remove it from our systems within 30 days unless we need to retain it for longer as described below, or it has already been de-identified and disassociated from your account when you allow us to use your Content to”
Human review of chats
Authorized OpenAI staff and contractors may access content for abuse/security investigations, support, legal matters, and model improvement (the last one stops if you opt out).
- help.openai.com/en/articles/7039943-how-openai-handles-data-in-consumer-services · checked Oct 17, 2026
“A limited number of authorized OpenAI personnel, as well as trusted service providers that are subject to confidentiality and security obligations, may access user content only as needed for these reasons: (1) investigating abuse or a security incident; (2) to provide support to you if you reach out to us with questions about your account, or to troubleshoot issues affecting the product experience; (3) to handle legal matters; or (4) to improve model performance (unless you have opted out).”
- help.openai.com/en/articles/7039943-how-openai-handles-data-in-consumer-services · checked Oct 17, 2026
“Please do not enter sensitive information that you would not want reviewed or used.”
Temporary / incognito chat
New chat > Temporary. Not in history, not used for training while temporary, copy kept up to 30 days for safety. Saving it turns it into a regular chat under your normal settings.
- help.openai.com/en/articles/8914046-temporary-chat-faq · checked Oct 17, 2026
“A temporary chat is not used to improve OpenAI models while it remains temporary, whether or not you choose personalization.”
- help.openai.com/en/articles/8914046-temporary-chat-faq · checked Oct 17, 2026
“OpenAI may keep a copy of a temporary chat for up to 30 days for safety purposes.”
Data location
Consumer content is stored in the US and elsewhere. The help pages describe data residency only for business workspaces.
- help.openai.com/en/articles/7039943-how-openai-handles-data-in-consumer-services · checked Oct 17, 2026
“Content is stored on OpenAI systems and our trusted service providers’ systems in the US and around the world.”
- openai.com/policies/us-privacy-policy/ · checked Oct 17, 2026
“including processing and storing your Personal Data in our facilities and servers in the United States, or in countries or territories where our affiliates and partners or our vendors and service providers are located.”
SOC 2 Type II
Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed).
ISO/IEC 27001
Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed).
ISO/IEC 42001
Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed).
DPA available
Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed).
BAA / HIPAA
Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed).
Zero-retention option
Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed).
How to turn off training in ChatGPT (Free / Plus / Pro)
Turn off "Improve the model for everyone".
Web
- Open your account menu.
- Select Settings.
- Select Data controls.
- Select Improve the model for everyone, turn it off, and select Done.
iOS and Android
- Open the sidebar.
- Select your profile icon to open Settings.
- Select Data controls.
- Turn off Improve the model for everyone.
Alternative
Or select "Do not train on my content" in the OpenAI Privacy Portal (either option is sufficient).
Keep in mind
- Applies to new conversations only.
- Thumbs up/down feedback can still send the whole conversation to training.
- Codex has a separate "Include environments" setting.
- help.openai.com/en/articles/7730893-data-controls-faq · checked Oct 17, 2026
“Select Improve the model for everyone, turn it off, and select Done.”
- help.openai.com/en/articles/7730893-data-controls-faq · checked Oct 17, 2026
“Turn off Improve the model for everyone.”
Vendor pages we monitor for this tool
We save a fingerprint (SHA-256 of the page's visible text) on every check so we can spot when a vendor edits its terms.
| Page | Checked | Fingerprint |
|---|---|---|
| https://help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance | Oct 17, 2026 | 799593234b10… |
| https://help.openai.com/en/articles/7730893-data-controls-faq | Oct 17, 2026 | 872ec5063ce1… |
| https://help.openai.com/en/articles/7039943-how-openai-handles-data-in-consumer-services | Oct 17, 2026 | 475270ca9008… |
| https://help.openai.com/en/articles/8809935-deleting-and-archiving-chats-in-chatgpt | Oct 17, 2026 | 4e09bb99d1d3… |
| https://help.openai.com/en/articles/8914046-temporary-chat-faq | Oct 17, 2026 | 1ccd00143baf… |
| https://openai.com/policies/us-privacy-policy/ | Oct 17, 2026 | e44b9d4b6a3e… |
Compare with
- Claude (Free / Pro / Max) You choose at signupPersonal
- Gemini app (personal Google account) Trains by defaultPersonal
- Microsoft Copilot (personal Microsoft account) Depends on app versionPersonal
- Amazon Q Business Doesn't train
- ChatGPT Enterprise / Business Doesn't train
- Claude for Work (Team / Enterprise) Doesn't train
See all Assistant tools · Why AI training on your data matters · How we source and verify this