What is the right to be forgotten
Under some privacy laws, including the GDPR, you may be able to ask OpenAI to stop certain personal information about you from appearing in ChatGPT responses. This may apply where the information is inaccurate, excessive, irrelevant, or no longer appropriate. This type of request is sometimes referred to as an “erasure”, “right to be forgotten” or “objection” request. If you’d like to make one of these requests, you can submit a “Remove my personal data from ChatGPT responses” request via our Privacy Portal.
Requests are assessed on a case-by-case basis and involve balancing an individual's privacy rights against other important considerations, such as freedom of expression and the public interest (see How we review requests below).
Why ChatGPT shows information about me
As a result of learning language, ChatGPT responses may sometimes include personal information about individuals whose personal information appears multiple times on the public internet (for example, public figures). You can learn more about how ChatGPT works here: How ChatGPT and our foundation models are developed.
Depending on the question you ask, ChatGPT may also search the web and cite sources in its response to you, which might also result in the response containing personal information about you.
Who can make a request
Whether or not you have a ChatGPT account, you can ask OpenAI to prevent information about you from appearing in ChatGPT responses if you believe that it is inaccurate, excessive, irrelevant, or no longer appropriate.
Requests are typically submitted by the person concerned, but requests may also be made by a legally authorised representative acting on someone else’s behalf. In such cases, we may ask for documentation confirming that authority.
How to submit a request
Go to our Privacy Portal
Click Make a Privacy Request
Select Remove personal data from ChatGPT responses
Provide the following details:
Personal details that appear in ChatGPT responses (e.g. name, email, phone)
Links to specific ChatGPT chats referencing your personal details
Detailed reasons why you believe the information should be removed (i.e. precise reasons for the data being inaccurate, excessive, irrelevant or no longer appropriate)
Proof of identity (e.g. government-issued ID)
Contact email
Submit the request
We may contact you for clarification or additional information if needed.
Best practices for a successful request
Before submitting a request, it is helpful to:
Be specific. Provide specific examples of where the information appears, including links to chats, and URLs referenced in these chats.
Identify the type of data. Indicate whether it’s a name, email, phone number, or specific URL so our team can locate it quickly.
Explain why removal is necessary. Describe why the data is inaccurate, excessive, irrelevant, or no longer appropriate.
Prepare verification documents. A government-issued ID or alternative proof will be required so be ready to submit it.
Providing specific examples helps us assess requests more accurately and efficiently.
How we review requests
When reviewing a request, we assess the specific circumstances of your request in light of applicable privacy laws. Each request involves a contextual balancing of the individual’s privacy rights against other relevant interests, including freedom of expression and the public interest.
In particular, we may take into account factors such as:
Role in public life
We consider whether the person is a private individual or has a public or professional role, and whether the information relates to that role. Information closely connected to a person’s public or professional activities may be more likely to remain relevant.Accuracy and truthfulness
We consider whether the information is inaccurate, misleading, incomplete, or omits important context. Where you believe information is incorrect, you should explain why and provide reliable supporting evidence. We assess accuracy based on the information provided with your request and cannot independently investigate disputed facts. Information that is clearly incorrect or outdated (especially where authoritative sources, such as a court decision, clarify the position) may weigh in favour of removal.Ongoing relevance
We assess how old the information is, including whether the information continues to reflect the individual’s current situation. Information about events that occurred long ago may be less relevant over time, particularly where circumstances have changed.Nature and sensitivity of the information
We look at whether the information involves particularly sensitive personal information (for example, information about health or religion), and the potential impact its continued appearance may have on an individual’s private life. Sensitive information may carry greater weight in favour of removal, particularly where it is no longer relevant or was not intentionally made public.Public interest
We assess whether there is a legitimate public interest in the information remaining available in responses generated by ChatGPT. This includes considering whether the information contributes to reporting on matters of public importance, informs public debate, or helps users understand or evaluate a person’s public or professional role or conduct including where the information may be relevant to users in assessing professional services and expertise.Impact on the individual
We consider the potential effects of the information continuing to appear in ChatGPT responses. This includes considering whether the information has caused, or could reasonably be expected to cause, unfair or ongoing effects in light of the overall context, such as adverse impacts on private life, safety or future opportunities.Official sources
We consider whether the information originates from public authorities, official registers, or other sources performing a public task, and whether continued access remains necessary or relevant. While such information may carry public interest, this does not automatically mean it should remain available indefinitely, particularly where circumstances have changed.Criminal convictions, allegations, or misconduct
We consider whether the information relates to criminal convictions, allegations, or other forms of misconduct, and the extent to which it remains relevant. This may include considering whether the information reflects the outcome of proceedings, whether it is accurate and up to date, how much time has passed, and whether there is an ongoing public interest in the information remaining available.Children
We consider whether the information relates to a person who was a child at the time the information was created or published. Information about children or young people generally attracts stronger privacy interests.
No single factor is decisive as we consider all relevant factors together. A single, heavily weighted factor may outweigh several factors pointing the other way.
Where necessary, we may ask for additional information to help us understand the context of a request and complete our assessment.
FAQ
Will my request be always approved?
Not necessarily. In accordance with privacy laws, “erasure”, “right to be forgotten” or “objection” rights are not absolute and we may decline a request if we have a lawful reason for doing so. If you feel we have not adequately addressed your request, you have the right to lodge a complaint with your local supervisory authority. Please see our Privacy Policy for more information.
Will OpenAI also remove the information from search engines?
No. Removing your personal data from ChatGPT prevents it from appearing in ChatGPT responses but does not remove it from external websites or search engines. To do this, you will need to contact other platforms separately.