AI Policy Log

Tool page

OpenAI / ChatGPT

Official OpenAI pages distinguish consumer ChatGPT (Free, Plus, Pro) from business workspaces and the API. Consumer data sharing is enabled by default and can be turned off. Business, Enterprise, Edu, Healthcare, Teachers, and API inputs are not used for training by default. This is a quotation log, not a legal opinion.

Checked 2026-11-18 (America/Toronto). Page status: complete · Vendor: OpenAI

Not a verdict. These are quotations from official vendor pages. They are not legal advice and not a determination that a plan is safe, compliant, or suitable for a particular kind of data.

Policy by plan

Plan Trains on inputs by default? Opt-out / opt-in Retention ZDR BAA / DPA Data residency Exceptions
ChatGPT Free, Plus, and Pro (services for individuals) Yes, for eligible content, unless you opt out. Official help text: “On personal workspaces with ChatGPT Free, Plus, or Pro, data sharing is enabled by default.” Turn off “Improve the model for everyone” under Settings > Data controls, or select “Do not train on my content” in the Privacy Portal. Either option is described as sufficient for ChatGPT conversations and Codex tasks. Cleared chats are deleted from OpenAI’s systems within 30 days, unless previously de-identified and disassociated from the account or retained for security or legal reasons. A numbered retention period for ordinary saved chat history was not stated / not verified on the pages we read. Temporary chats may be retained for up to 30 days for safety purposes. Not stated / not verified as a consumer ChatGPT feature on the pages we read. Zero Data Retention is described for eligible API endpoints. A DPA is described for ChatGPT Business, ChatGPT Enterprise, and the API, not for these consumer plans, on the Enterprise privacy page. Data residency and inference residency are described for eligible API customers and new ChatGPT Enterprise/Edu customers, not as a consumer-plan control, on the residency help article. Even after opt-out, thumbs-up/down feedback may allow the entire associated conversation to be used for training. Temporary chats are not used to improve models while they remain temporary. Trusted Access programs such as Daybreak Blue may use session content to improve models even when “Improve the model for everyone” is off. Voice mode has a separate training control. OpenAI Support conversations may be used to improve models if the setting is enabled.
ChatGPT Business No, by default. “By default, we do not use your business data for training our models.” The same default is stated for ChatGPT Business, Enterprise, Edu, Healthcare, Teachers, and the API. Training requires an explicit opt-in or a separate agreement. Response feedback (thumbs up/down) “does not permit training on the associated conversation.” Workspace admins can enable or disable response feedback. Workspace admins can control how long data is retained. “Any deleted or unsaved conversations are removed from our systems within 30 days, unless longer retention is required by law, or is reasonably necessary to protect our services or any third party from harm.” Not stated / not verified as a ChatGPT Business workspace feature on the pages we read. OpenAI says it can execute a DPA with customers for ChatGPT Business, ChatGPT Enterprise, and the API. The residency help article describes eligibility for eligible API customers and new ChatGPT Enterprise/Edu customers. Business-plan residency was not stated / not verified there. By default, OpenAI does not train on data accessed from apps. Public GPT sharing, if enabled by admins, may be subject to additional review.
ChatGPT Enterprise, ChatGPT Edu, ChatGPT for Healthcare, and ChatGPT for Teachers No, by default. Enterprise privacy: “We do not train our models on your data by default,” covering business data from ChatGPT Business, Enterprise, Healthcare, Edu, Teachers, and the API Platform. Not used for training unless the organization has signed a separate agreement allowing it. For Enterprise, Business, and Edu, thumbs feedback does not permit training on the associated conversation. For Enterprise, Edu, and Healthcare: “Your workspace admins control how long your data is retained. Any deleted conversations are removed from our systems within 30 days, unless we are legally required to retain them.” For Teachers: “Each of your workspace members control whether their conversations are retained,” with the same 30-day deletion rule unless legally required. Not stated / not verified as a standard ChatGPT Enterprise/Edu/Healthcare feature on the pages we read. ZDR is described for eligible API endpoints. DPA available for Business, Enterprise, and the API. For Edu and Teachers, OpenAI says it processes data under a Student Data Privacy Agreement. HIPAA-eligible products listed with a BAA include ChatGPT for Healthcare, ChatGPT for Enterprise with Regulated Workspace, ChatGPT FedRAMP, ChatGPT for Clinicians, and certain API offerings. Some features (including improved memory) are described as not covered under the BAA. Eligible new ChatGPT Enterprise/Edu customers can store in-scope content at rest in listed regions (including Australia, Canada, Europe (EEA + Switzerland), India, Japan, Singapore, South Korea, UAE, UK, and United States, as of the article we read). Inference residency is described for Europe (EEA + Switzerland), the United States, and the UAE, with stated feature limits. Data residency is included at no additional cost for ChatGPT Enterprise and Education plans, per that article. Authorized employee access is described as limited (incidents, recovery with permission, or law). Some ChatGPT functionality is explicitly “not covered under the BAA”; PHI should not be entered there. Live third-party search is distinguished from OpenAI’s own search index in HIPAA-eligible workspaces.
API Platform No, by default after March 1, 2023, unless the organization opts in. “As of March 1, 2023, data sent to the OpenAI API is not used to train or improve OpenAI models (unless you explicitly opt in to share data with us).” Organization owners can opt in, at organization or selected-project level, to share feedback, evaluation and fine-tuning data, and/or API inputs and outputs. “Accounts with Zero Data Retention enabled cannot opt in to our data sharing mechanisms.” Input/output sharing “is not available to certain organizations, including Enterprise and customers with Zero Data Retention enabled.” “Except for certain endpoints and features listed in our platform documentation, OpenAI may securely retain API inputs and outputs for up to 30 days to provide the services and to identify abuse.” Fine-tuning data is retained until the customer deletes the files. Eligible customers may request Zero Data Retention for eligible endpoints, subject to prior approval and additional requirements. ZDR “excludes customer content from abuse monitoring logs.” An August 19, 2026 post (updated September 22, 2026) describes Private Safety Processing as compatible with ZDR. CSAM-flagged images “will continue to be retained for manual review and reporting purposes, even in Zero Data Retention deployments.” DPA available. OpenAI says it can sign a BAA; “Please reach out if you require a BAA.” HIPAA eligibility for the API is described as requiring an executed BAA and Modified Retention unless OpenAI specifies otherwise, and applies to listed endpoints. Eligible API customers can choose to have in-scope customer content stored at rest in supported countries, per the ChatGPT/API residency documentation. Details should be read on the current API data-residency guide. Endpoint eligibility for ZDR varies; some endpoints may still store application state. ZDR with Private Safety Processing has additional configuration requirements.

Source quotes

“On personal workspaces with ChatGPT Free, Plus, or Pro, data sharing is enabled by default. You can turn off Improve the model for everyone and keep your chat history.” ChatGPT Free, Plus, and Pro (services for individuals) · Checked 2026-11-18 · Data controls in ChatGPT (OpenAI Help Center)
“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, or select Do not train on my content in our Privacy Portal.” Consumer training and opt-out · Checked 2026-11-18 · How your data is used to improve model performance
“For individual plans (Free, Plus, and Pro), if you submit feedback on a response—for example, by selecting thumbs up or thumbs down—we may use the entire conversation associated with that feedback to improve our models, even if you have opted out of model training.” Feedback exception · Checked 2026-11-18 · How your data is used to improve model performance
“While they remain temporary, these chats: do not appear in your chat history; do not create or update memories; are not used to improve OpenAI models; may be retained for up to 30 days for safety purposes.” Temporary chats · Checked 2026-11-18 · Data controls in ChatGPT
“During Daybreak sessions on your personal account, we may use content from your interactions to improve our models even when “Improve the model for everyone” is off.” Trusted Access / Daybreak Blue · Checked 2026-11-18 · Data controls in ChatGPT
“Cleared chats are deleted from our systems within 30 days, unless they have previously been de-identified and disassociated from your account or we have to keep them for security or legal reasons.” Deletion timing · Checked 2026-11-18 · How OpenAI handles data in consumer services
“By default, we do not use your business data for training our models. If you have explicitly opted in to share your data with us (for example, through our opt-in feedback mechanisms) to improve our services, then we may use the shared data to train our models.” ChatGPT Business · Checked 2026-11-18 · Enterprise privacy at OpenAI (updated January 8, 2026)
“By default, data from ChatGPT Business, ChatGPT Enterprise, ChatGPT for Healthcare, ChatGPT Edu, ChatGPT for Teachers, and the API Platform (after March 1, 2023) isn’t used for training our models, unless you have explicitly opted in to share your data with us to improve the services.” ChatGPT Enterprise, ChatGPT Edu, ChatGPT for Healthcare, and ChatGPT for Teachers · Checked 2026-11-18 · Enterprise privacy at OpenAI (updated January 8, 2026)
“As of March 1, 2023, data sent to the OpenAI API is not used to train or improve OpenAI models (unless you explicitly opt in to share data with us).” API Platform · Checked 2026-11-18 · Your data — OpenAI API platform docs

Sources checked

Changelog for this tool

No approved later changes. First compiled 2026-11-18.