OpenAI
OpenAI's terms give the company sweeping power over your data and your legal rights, with your conversations used to train its AI models by default, deleted data potentially persisting forever inside those models, and your personal information shared with marketing partners and governments at OpenAI's discretion. If something goes wrong, your ability to fight back is severely limited — US users must take disputes to private arbitration with no class-action option, and OpenAI's financial liability to you is capped at just $100 regardless of the harm caused. OpenAI can change both its Terms of Service and Privacy Policy at any time without your direct agreement, and simply continuing to use the service counts as acceptance. There are some genuine user protections — you own your outputs, data export and training opt-out tools exist, and EEA and UK users keep their court rights — but the overall balance of power strongly favors OpenAI.
Last analyzed
Heads up: this analysis is AI-generated, can be incomplete or wrong, and is not legal advice. Use it as a starting point and read the original document before you agree — see our Terms.
Clause-by-clause breakdown
Every clause we flagged in OpenAI's agreement, with what it means for you.
All disputes must go to private arbitration, not court.
“MANDATORY ARBITRATION. You and OpenAI agree to resolve any claims arising out of or relating to these Terms or our Services, regardless of when the claim arose, even if it was before these Terms existed (a "Dispute"), through final and binding arbitration. You may opt out of arbitration within 30 days of account creation or of any updates to these arbitration terms within 30 days after the update has taken effect by filling out this form.”
Quoted verbatim from the agreement What this means: By default, you cannot sue OpenAI in court — all disputes must go through private arbitration. This applies even to claims that arose before you agreed to these Terms. You have a 30-day window from account creation (or from when an update takes effect) to opt out by submitting a form; if you miss that window, you lose your right to sue in court.
What you can do: If you created your account recently or OpenAI has updated its arbitration terms, submit the opt-out form at openai.com/form/arbitration-opt-out within 30 days. Keep a record of your submission.
You cannot join class-action lawsuits or get a jury trial.
“CLASS AND JURY TRIAL WAIVERS. You and OpenAI agree that Disputes must be brought on an individual basis only, and may not be brought as a plaintiff or class member in any purported class, consolidated, or representative proceeding. Class arbitrations, class actions, and representative actions are prohibited. Only individual relief is available. The parties agree to sever and litigate in court any request for public injunctive relief after completing arbitration for the underlying claim and all other claims. This does not prevent either party from participating in a class-wide settlement. You and OpenAI knowingly and irrevocably waive any right to trial by jury in any action, proceeding, or counterclaim.”
Quoted verbatim from the agreement What this means: You cannot join a class-action lawsuit against OpenAI, even if many users are harmed in the same way. You also permanently give up your right to a jury trial. This means if OpenAI harms thousands of users in a small but identical way, each person must fight their case alone — making it economically impractical to pursue most claims.
What you can do: Opting out of arbitration (see the mandatory arbitration finding) also removes the class-action waiver. Act within 30 days of account creation or a terms update.
OpenAI's maximum liability to you is $100 or your last 12 months of payments.
“NEITHER WE NOR ANY OF OUR AFFILIATES OR LICENSORS WILL BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR EXEMPLARY DAMAGES, INCLUDING DAMAGES FOR LOSS OF PROFITS, GOODWILL, USE, OR DATA OR OTHER LOSSES, EVEN IF WE HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. OUR AGGREGATE LIABILITY UNDER THESE TERMS WILL NOT EXCEED THE GREATER OF THE AMOUNT YOU PAID FOR THE SERVICE THAT GAVE RISE TO THE CLAIM DURING THE 12 MONTHS BEFORE THE LIABILITY AROSE OR ONE HUNDRED DOLLARS ($100).”
Quoted verbatim from the agreement What this means: Even if OpenAI's service causes you significant harm — such as data loss, financial damage, or reliance on incorrect AI output — the most you can ever recover is either what you paid in the past 12 months or $100, whichever is greater. All indirect or consequential damages (like lost profits or lost data) are completely excluded. For free-tier users, the cap is just $100.
What you can do: Do not rely on ChatGPT for high-stakes decisions (financial, medical, legal) where a mistake could cost you more than $100. Always verify important outputs independently.
OpenAI can change the Terms; continued use means you agree.
“We will give you at least 30 days advance notice of changes to these Terms that materially adversely impact you either via email or an in-product notification. All other changes will be effective as soon as we post them to our website. If you do not agree to the changes, you must stop using our Services.”
Quoted verbatim from the agreement What this means: OpenAI can change these Terms at any time. For material changes, they'll give 30 days' notice, but for anything they don't consider materially adverse, changes go live the moment they're posted. Continuing to use the service after a change means you've accepted the new terms — even if you didn't read them.
What you can do: Watch for email notifications from OpenAI about Terms changes. If you disagree with a change, stop using the service before the effective date to avoid being bound by the new terms.
OpenAI can change this policy anytime without your direct consent.
“We may update this policy from time to time. When we do, we will publish an updated version and effective date on this page, unless another type of notice is required by applicable law.”
Quoted verbatim from the agreement What this means: OpenAI can change this Privacy Policy at any time simply by posting a new version on their website. There is no requirement to notify you directly or to get your affirmative consent before the new terms take effect. Continued use of the service after a change implicitly binds you to the new terms.
What you can do: Periodically check the Privacy Policy page for updates. Consider setting a reminder to review it every few months.
Your data is shared with marketing partners for targeted advertising.
“We also share limited information with select marketing partners who are not service providers in order to promote our products and services on third-party properties and help us assess the effectiveness of those efforts. Some of these partners may receive information through cookies and similar technologies.”
Quoted verbatim from the agreement What this means: OpenAI shares some of your data with marketing partners — companies that are not just service providers — to run targeted ads for OpenAI's own products on other websites. These partners can receive data via cookies and tracking technologies. This is classified as 'cross-context behavioral advertising' under some state laws.
What you can do: Opt out via the marketing privacy control in your account settings, or use the 'Manage Cookies' link on the website if you're not logged in. You can also use Global Privacy Control in your browser.
Deleted data may persist in AI models if used for training.
“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 improve our models.”
Quoted verbatim from the agreement What this means: When you delete your data, OpenAI has up to 30 days to actually remove it. More importantly, if you previously allowed your content to be used for model training, that content may have already been de-identified and incorporated into the model — meaning it effectively cannot be deleted even if you request it.
What you can do: Opt out of model training before generating sensitive content. Once content has been de-identified and used for training, deletion requests will not remove it from the model.
OpenAI can use your feedback freely with no payment to you.
“Feedback. We appreciate your feedback, and you agree that we may use it without restriction or compensation to you.”
Quoted verbatim from the agreement What this means: Any feedback, suggestions, or ideas you share with OpenAI — such as feature requests or bug reports — become OpenAI's property to use however they like, with no obligation to credit or pay you. This is a broad, irrevocable grant.
What you can do: Be aware that any suggestions you submit to OpenAI can be used commercially without compensation. This is common industry practice but worth knowing.
Your employer can take over and read your account if you used a work email.
“Corporate domains. If you create an account using an email address owned by an organization (for example, your employer), that account may be added to the organization's business account with us, in which case we will provide notice to you so that you can help facilitate the transfer of your account (unless your organization has already provided notice to you that it may monitor and control your account). Once your account is transferred, the organization's administrator will be able to control your account, including being able to access Content (defined below) and restrict or remove your access to the account.”
Quoted verbatim from the agreement What this means: If you signed up with a work email address, your employer can claim control of your account. Once that happens, your employer can read all your chat history (Content), restrict what you can do, or even remove your access entirely. This could expose personal conversations you had before the transfer.
What you can do: Use a personal email address to create your OpenAI account if you want to keep your conversations private from your employer. If you already used a work email, consider what content you've shared.
Paid subscriptions are non-refundable when you cancel.
“Payments are non-refundable, except where required by law. These Terms do not override any mandatory local laws regarding your cancellation rights.”
Quoted verbatim from the agreement What this means: If you pay for a ChatGPT Plus or other subscription and cancel mid-cycle, you will not receive a refund for the unused portion of your subscription period. The only exceptions are where local consumer protection laws require refunds.
What you can do: Cancel your subscription close to the renewal date to avoid paying for unused time. Check your local consumer protection laws — in some regions (e.g., EU) you may have stronger refund rights.
OpenAI makes no guarantees about service quality or data safety.
“OUR SERVICES ARE PROVIDED "AS IS." EXCEPT TO THE EXTENT PROHIBITED BY LAW, WE AND OUR AFFILIATES AND LICENSORS MAKE NO WARRANTIES (EXPRESS, IMPLIED, STATUTORY OR OTHERWISE) WITH RESPECT TO THE SERVICES, AND DISCLAIM ALL WARRANTIES INCLUDING, BUT NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, SATISFACTORY QUALITY, NON-INFRINGEMENT, AND QUIET ENJOYMENT, AND ANY WARRANTIES ARISING OUT OF ANY COURSE OF DEALING OR TRADE USAGE. WE DO NOT WARRANT THAT THE SERVICES WILL BE UNINTERRUPTED, ACCURATE OR ERROR FREE, OR THAT ANY CONTENT WILL BE SECURE OR NOT LOST OR ALTERED.”
Quoted verbatim from the agreement What this means: OpenAI makes no promises about the quality, accuracy, or reliability of its services. They explicitly disclaim any guarantee that the service will work correctly, be available, or keep your data safe. If the AI gives you wrong information or your data is lost, OpenAI has made no warranty that this won't happen.
What you can do: Always independently verify important information from ChatGPT. Do not use it as the sole source for medical, legal, financial, or other high-stakes decisions.
Your data can be sold or transferred if OpenAI is acquired.
“If we are involved in strategic transactions, reorganization, bankruptcy, receivership, or transition of service to another provider (collectively, a "Transaction"), your Personal Data may be disclosed in the diligence process with counterparties and others assisting with the Transaction and transferred to a successor or affiliate as part of that Transaction along with other assets.”
Quoted verbatim from the agreement What this means: If OpenAI is sold, merges with another company, or goes bankrupt, all of your personal data can be transferred to the new owner as a business asset. You have no right to object to this transfer, and the new owner may have different privacy practices.
What you can do: There is no way to prevent this if you use the service. Deleting your account before any such transaction is announced is the only way to reduce exposure.
Connecting contacts uploads data about people who don't use OpenAI.
“If you choose to connect your device contacts, we upload information from your device address books and check which of your contacts also use our Services. If any of your contacts aren't yet using our Services, we'll update you if they sign up for our Services later.”
Quoted verbatim from the agreement What this means: If you grant OpenAI access to your phone contacts, it uploads your entire address book — including data about people who have never agreed to OpenAI's terms. OpenAI then tracks when those non-users sign up and notifies you. These third parties have no say in this data collection.
What you can do: Do not grant OpenAI access to your device contacts unless you are comfortable with your contacts' information being uploaded and stored by OpenAI.
OpenAI builds a detailed profile of your behavior across its services.
“We collect information about your use and activity across the Services, such as the types of content that you view or engage with, the features you use and the actions you take, when you submit feedback to a model response, the people with whom you interact, as well as your time zone, country, the dates and times of access, user agent and version, type of computer or mobile device, and your computer connection.”
Quoted verbatim from the agreement What this means: OpenAI collects a detailed behavioral profile of how you use its services, including what content you view, what features you use, who you interact with, and your device and connection details. This is a broad dataset that goes well beyond what is needed to simply answer your questions.
What you can do: Review your account's data controls settings to limit what is retained. Using Temporary Chat mode reduces some data collection.
OpenAI scrapes public internet data to train its AI models.
“We also collect information from other sources, like information that is publicly available on the internet, to develop the models that power our Services.”
Quoted verbatim from the agreement What this means: OpenAI scrapes publicly available internet data to train its models. This means your publicly posted content — blog posts, social media, forum comments — may have been used to train ChatGPT without your direct knowledge or consent.
What you can do: Visit the help center article linked in the policy for information on how to exercise rights regarding data collected from the internet for training purposes.
You can opt out of having your data used for AI training.
“Opt out. If you do not want us to use your Content to train our models, you can opt out by following the instructions in this article. Please note that in some cases this may limit the ability of our Services to better address your specific use case.”
Quoted verbatim from the agreement What this means: OpenAI does provide a way to opt out of having your content used for model training, which is a meaningful user protection. However, the opt-out is not the default — you must actively find and complete the process. The warning that opting out may limit service quality could discourage users from exercising this right.
What you can do: Visit openai.com/policies/how-your-data-is-used-to-improve-model-performance and follow the opt-out instructions to prevent your conversations from being used in training.
You own both your inputs and the AI-generated outputs.
“Ownership of content. As between you and OpenAI, and to the extent permitted by applicable law, you (a) retain your ownership rights in Input and (b) own the Output. We hereby assign to you all our right, title, and interest, if any, in and to Output.”
Quoted verbatim from the agreement What this means: OpenAI explicitly assigns ownership of AI-generated outputs to you. You also retain ownership of what you type in. This is a user-friendly provision that means you can use, publish, or commercialize the content generated for you.
What you can do: Note that outputs may not be unique — other users can receive similar content — so copyright protection for AI outputs may be limited under applicable law.
OpenAI will refund prepaid amounts if they shut down a service.
“We may decide to discontinue our Services, but if we do, we will give you advance notice and a refund for any prepaid, unused Services.”
Quoted verbatim from the agreement What this means: If OpenAI shuts down a service you've paid for, they commit to giving you advance warning and refunding any prepaid amount you haven't used. This is a user-protective commitment.
You can export all your ChatGPT data directly from settings.
“You can export your ChatGPT history and data in your account's data controls.”
Quoted verbatim from the agreement What this means: OpenAI provides a self-service data export feature, allowing you to download your ChatGPT conversation history and account data. This is a user-protective feature that goes beyond the minimum required in many jurisdictions.
What you can do: Go to Settings → Data Controls → Export Data to download a copy of your ChatGPT history at any time.
Frequently asked questions
Is OpenAI's Terms & Conditions agreement fair?
- BeforeYouClick grades OpenAI E on an A–E scale, where A is the fairest. Its agreement contains 7 critical clauses, 8 warnings and 4 good practices.
What are the biggest problems with OpenAI's terms?
- The most serious clauses we flagged are: Forced Arbitration for All Disputes, Class Action and Jury Trial Waived and Liability Capped at $100 for All Harm.
How long does it take to read OpenAI's terms?
- OpenAI's agreement takes about 13 minutes to read in full, across 19 clauses we reviewed. The summary on this page covers the same ground in under a minute.
When was OpenAI's agreement last analyzed?
- We last read OpenAI's Terms & Conditions on August 4, 2026. Companies change their terms without notice, so check the original document for the current version.
Is this analysis legal advice?
- No. This analysis is AI-generated, can be incomplete or wrong, and is not legal advice. Use it as a starting point and read the original agreement before you agree to it.
More agreements to read
- AnthropicAIGrade E
- PosteoProductivityGrade A
- Mullvad VPNSecurityGrade A
- MurenaSecurityGrade B
- TutaProductivityGrade B
- NordVPNSecurityGrade B