Privacy Policy

Use & Processing · Context

The policy covers personal data from your use of OpenAI's website, apps and services

This US privacy policy describes OpenAI's practices for personal data it collects from or about you, and how it uses that data, when you use its website, applications and services, which the policy calls the "Services". "OpenAI", "we", "our" and "us" include OpenAI's affiliates, so the policy speaks for the group and not only one company. The opening text names ChatGPT as an example tool but defines Services broadly.

Supporting quotations

This Privacy Policy describes our practices with respect to personal data that we collect from or about you, and how we use it when you use our website, applications, and services (collectively, “Services”).

We at OpenAI (together with our affiliates, “OpenAI”, “we”, “our” or “us”)

Qualifications & exceptions

• "OpenAI", "we", "our" and "us" cover OpenAI together with its affiliates.
• "Services" collectively means the website, applications and services.

Not stated here

• The policy does not list which specific products count as Services beyond the examples given.

Cross-references

• has exception (model-proposed): Content processed for business customers, such as the API, is outside this policy — The policy's general scope is personal data from your use of the Services; content processed on behalf of business customers, such as the API, is excluded and governed by customer agreements.

Use & Processing · Context

Training details and choices are in a separate help article, plus statements in this policy

For how OpenAI collects and uses training information to develop its language models, and the choices people have, the policy points readers to this policy and to a linked help center article, which is not supplied here. This policy itself says OpenAI collects publicly available internet information to develop the models that power its Services, and that it may use Content you provide to improve the Services, for example to train the models that power ChatGPT, with linked instructions for opting out of that use of your Content.

Supporting quotations

For information about how we collect and use training information to develop our language models that power ChatGPT and other Services, and your choices with respect to that information, please see this policy as well as

We also collect information from other sources, like information that is publicly available on the internet, to develop the models that power our Services.

we may use Content you provide us to improve our Services, for example to train the models that power ChatGPT

on how you can opt out of our use of your Content to train our models

Qualifications & exceptions

• The opt-out described applies to OpenAI's use of your Content to train its models, per the linked instructions.
• Information collected from the public internet for model development is separate from Content you provide.

Not stated here

• The help center articles are linked but not supplied, so the detail of training sources and the opt-out steps is unknown.
• The policy does not say here whether the opt-out covers publicly collected internet information.

Use & Processing · Company practice

Data from advertisers, partners and the public internet is used to measure effectiveness and develop models

Information OpenAI receives from advertisers and other data partners is used for the purposes described in the policy, including to measure and improve the effectiveness of the Services, including ads shown to Free and Go users. Separately, OpenAI says it collects information from other sources such as publicly available internet information to develop the models that power its Services; a linked help article gives more on the sources used for ChatGPT and other Services.

Supporting quotations

which we use for purposes described in this Policy including to help us measure and improve the effectiveness of our Services, including ads shown to Free and Go users on our Services.

We also collect information from other sources, like information that is publicly available on the internet, to develop the models that power our Services.

Qualifications & exceptions

• The ad-effectiveness example is stated for ads shown to Free and Go users.
• Public internet information is described as being used to develop models.

Not stated here

• The help article on model-development sources is linked but not supplied.
• The policy does not list which advertisers or data partners provide data.

Use & Processing · Company practice

OpenAI uses personal data to run, improve and personalize the Services and for research

OpenAI lists these purposes for using Personal Data: to provide, analyse and maintain the Services (for example answering your questions in ChatGPT), to improve and develop the Services and conduct research (for example developing new features), and to personalize and customize your experience (for example giving you more relevant Content). These are stated as examples of broader purposes.

Supporting quotations

We use Personal Data for the following purposes:

To provide, analyse, and maintain our Services, for example to respond to your questions for ChatGPT;

To improve and develop our Services and conduct research, for example to develop new features;

To personalize and customize your experience across our Services, for example to provide you with more relevant Content;

Qualifications & exceptions

• Each purpose is given with an example, not an exhaustive scope.

Not stated here

• The policy does not say which data types are used for which purpose here; the state disclosures table repeats the purposes at a summary level.

Use & Processing · Company practice

For Free and Go users, personal data is used to personalize and measure ads, subject to your settings

For Free and Go users, OpenAI uses personal data to personalize the ads they see on the Services, subject to their settings, and to measure how effective the ads shown on the Services are. This ties to the earlier statement that, for Free and Go users, OpenAI collects information about their ads history and interests. A linked help article gives more on ads in ChatGPT.

Supporting quotations

For Free and Go users, to personalize the ads you see on our Services (subject to your settings), and to measure the effectiveness of ads shown on our Services.

users, we also collect information about your ads history and interests, such as the types of ad content that you view or engage with.

Qualifications & exceptions

• Stated for Free and Go users.
• Ad personalization is subject to your settings.

Not stated here

• The help article on ads in ChatGPT is linked but not supplied; the available settings are not described here.

Use & Processing · Company practice

OpenAI uses personal data to communicate with you, including information about the Services and events

OpenAI uses personal data to communicate with you, including to respond to your questions and to send you information about its Services and events, for example about changes or improvements to the Services.

Supporting quotations

To communicate with you, including to respond to your questions, and send you information about our Services and events, for example about changes or improvements to the Services;

Qualifications & exceptions

• Examples are not exhaustive.

Not stated here

• How to stop these communications is covered in the data controls section, not here.

Use & Processing · Company practice

Connected contacts are used to identify which of your contacts use the Services and to update you

When you choose to connect your contacts, OpenAI uses that information to identify which of your contacts use the Services and to update you if they join later.

Supporting quotations

Identify your contacts who use our Services when you choose to connect your contacts and update you if they join our Services later;

Qualifications & exceptions

• Only when you choose to connect your contacts.

Use & Processing · Company practice

Personal data and submitted Content are used to prevent fraud and misuse, including by monitoring Content

OpenAI uses personal data to prevent fraud, illegal activity or misuse of the Services and to protect the security of its systems and Services. The policy says this includes monitoring any Content submitted or exchanged on its platforms, so prompts and other Content you provide may be reviewed for these purposes. A linked page on transparency and content moderation gives more.

Supporting quotations

To prevent fraud, illegal activity, or misuses of our Services, and to protect the security of our systems and Services, including by monitoring any Content submitted or exchanged on our platforms

Qualifications & exceptions

• Monitoring is described as covering any Content submitted or exchanged on its platforms.

Not stated here

• The linked transparency and content moderation page is not supplied, so who reviews Content and how is unknown.

Use & Processing · Company practice

Personal data is used to meet legal duties, protect rights and safety, and estimate age

OpenAI uses personal data to comply with legal obligations and to protect the rights, privacy, safety or property of its users, OpenAI or third parties, for instance to prevent harm to you or others, or to estimate your age to give you an age-appropriate experience.

Supporting quotations

To comply with legal obligations and to protect the rights, privacy, safety, or property of our users, OpenAI, or third parties, for instance to prevent harm to you or others, or to estimate your age to give you an age-appropriate experience.

Qualifications & exceptions

• Age estimation is given as an example of this purpose.

Not stated here

• How age is estimated is not described in this passage.

Use & Processing · Company commitment

Aggregated or de-identified data is used for the same purposes and kept de-identified, unless required by law

OpenAI aggregates or de-identifies personal data so it no longer identifies you and uses that information for the purposes described above, such as analyzing how the Services are used, improving and adding features, and conducting research. It says it will maintain and use de-identified information in de-identified form and not try to re-identify it, unless required by law.

Supporting quotations

We also aggregate or de-identify Personal Data so that it no longer identifies you and use this information for the purposes described above, such as to analyze the way our Services are being used, to improve and add features to them, and to conduct research.

We will maintain and use de-identified information in de-identified form and not attempt to reidentify the information, unless required by law.

Qualifications & exceptions

• Commitment not to re-identify has an exception where required by law.

Not stated here

• The policy does not say what de-identification method is used.

Use & Processing · Permission or license

Your Content may be used to improve the Services, including training models, with an opt-out

OpenAI says it may use Content you provide to improve its Services, for example to train the models that power ChatGPT. It links instructions on how you can opt out of this use of your Content to train its models. The opening of the policy also refers readers to a help article for how training information is collected and used and the choices available, which is not supplied.

Supporting quotations

we may use Content you provide us to improve our Services, for example to train the models that power ChatGPT

on how you can opt out of our use of your Content to train our models

For information about how we collect and use training information to develop our language models that power ChatGPT and other Services, and your choices with respect to that information, please see this policy as well as

Qualifications & exceptions

• The opt-out is for OpenAI's use of your Content to train its models, per the linked instructions.
• The sentence says "may", not that all Content is always used.

Not stated here

• The linked instructions are not supplied, so the opt-out steps and any exceptions are unknown.
• The policy does not say here whether other data types besides Content are used for training.

Cross-references

• referred to by (model-proposed): You can choose whether your Content is used to improve and train models; Temporary Chats are not used — The data controls section's choice about training on your Content relates to the use section's statement that OpenAI may use Content to train models, with an opt-out.

Use & Processing · User right or choice

You can choose whether your Content is used to improve and train models; Temporary Chats are not used

You can choose whether your Content can be used to improve and train OpenAI's models, through the controls described in a linked help article. If you enable Temporary Chat in ChatGPT, those conversations will not appear in your history or be used to improve OpenAI's models. The policy's use section says OpenAI may use Content you provide to improve its Services, including to train models, and links instructions for opting out.

Supporting quotations

choose whether your Content can be used to improve and train our models

in ChatGPT, those conversations with ChatGPT will not appear in your history or be used to improve OpenAI’s models.

we may use Content you provide us to improve our Services, for example to train the models that power ChatGPT

We collect Personal Data that you provide in the input to our Services (“Content”), including your prompts and other content you upload, such as

Qualifications & exceptions

• Content here means the input you provide to the Services, such as prompts and uploads.
• Temporary Chat conversations are not used to improve models, but are deleted automatically within 30 days, unless kept for safety or legal reasons.

Not stated here

• The steps for the control are in a linked help article that is not supplied.
• The control is described for Content; the policy does not say here whether other data types are used for training.

Cross-references

• refers to (model-proposed): Your Content may be used to improve the Services, including training models, with an opt-out — The data controls section's choice about training on your Content relates to the use section's statement that OpenAI may use Content to train models, with an opt-out.
• defined by (model-proposed): Your prompts, uploads and interactions with other users are collected as Content — Content is defined as the input you provide to the Services; the training control applies to Content as defined.

Use & Processing · User right or choice

You can decide whether ChatGPT remembers details between chats

You can decide whether OpenAI remembers details between chats to make Content more personalized and relevant. The deletion section adds that you can delete specific Memories.

Supporting quotations

You can decide whether we will remember details between chats to make Content more personalized and relevant.

Qualifications & exceptions

• A personalization control; the passage describes the choice but not the default.

Not stated here

• The default setting and what is remembered are in a linked Memories help article that is not supplied.

Use & Processing · Company commitment

Data is processed on servers in various countries, including the US, under legally valid transfer mechanisms

OpenAI processes your Personal Data on servers in various jurisdictions, including in its facilities and servers in the United States and in countries where its affiliates, partners, vendors and service providers are located. It says it applies the protections in the policy to your Personal Data regardless of where it is processed and only transfers data under legally valid transfer mechanisms.

Supporting quotations

OpenAI processes your Personal Data for the purposes described in this policy on servers located in various jurisdictions, 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.

While data protection law varies by country, we apply the protections described in this policy to your Personal Data regardless of where it is processed, and only transfer that data pursuant to legally valid transfer mechanisms.

Qualifications & exceptions

• Transfers rely on "legally valid transfer mechanisms", which the policy does not name.

Not stated here

• The countries and the transfer mechanisms are not specified.

Use & Processing · Company commitment

OpenAI says it does not process sensitive data to infer characteristics about you

In the state disclosures introduction OpenAI states that it does not process sensitive data for the purpose of inferring characteristics about you. The statement is limited to that purpose; it does not say OpenAI never collects or processes sensitive data for other purposes.

Supporting quotations

We don’t process sensitive data for the purpose of inferring characteristics about you.

Qualifications & exceptions

• Limited to the purpose of inferring characteristics about you.

Not stated here

• The policy does not define "sensitive data" here or say whether sensitive data is collected for other purposes.

Use & Processing · Company practice

Table column 2: eight purposes for which this information is used, including training models subject to your control

The second column lists the purposes for which OpenAI says it uses this information: to provide, analyze and maintain the Services; to improve and develop the Services and conduct research, including using your Content to train models (subject to your control); to personalize your experience; for Free and Go users, subject to controls, to personalize and measure ads; to communicate with you, including about offers or information that may interest you; to identify your contacts who use the Services when you connect contacts; to prevent fraud and misuse, including by monitoring Content; and to comply with legal obligations and protect rights, including estimating age. The table's list has eight purposes. The earlier use section also lists promoting products through direct marketing and third-party properties as a separate purpose, which this column does not list as its own item; the "offers or information that may interest you" wording in the communications entry is the nearest match, and the opt-out text after the table addresses marketing sharing. This analysis does not read the omission as a limit on the earlier section.

Supporting quotations

We use this information for the following purposes, as described above:

Provide, analyze, and maintain our Services

Improve and develop our Services and conduct research, including using your Content to train our models (subject to your

To personalize and customize your experience across our Services

For Free and Go users, subject to your controls, to personalize the ads you see on our Services and measure the effectiveness of ads shown on our Services.

Communicate with you, including to respond to your questions, and send you information about our Services and events, for example about changes or improvements to the Services or offers or information that may interest you

To identify your contacts who use our Services when you choose to connect your contacts and update you if they join our Services later.

Prevent fraud, illegal activity, or misuses of our Services, and to protect the security of our systems and Services, including by monitoring any Content submitted or exchanged on our platforms

Comply with legal obligations and protect the rights, privacy, safety, or property of our users, OpenAI, or third parties, for instance to prevent harm to you or others, and to estimate your age to give you an age-appropriate experience

Qualifications & exceptions

• The ad purpose is for Free and Go users, subject to your controls.
• Using Content to train models is stated as subject to your control.
• The column is a list for the information as a whole, not matched to individual categories.

Not stated here

• Whether the table's omission of third-party-property promotion as its own purpose changes anything is not stated; the earlier section lists it and the opt-out text covers marketing sharing.

Cross-references

• contrast (model-proposed): OpenAI promotes its products through direct marketing and third-party properties, subject to your choices — The state disclosures table lists eight purposes and does not list promotion on third-party properties as its own item, while the use section lists it as a purpose subject to your choices. The text after the table covers marketing sharing.