Facebook is changing its privacy policy on June 26, 2024. This update means that public user content – excluding private messages – will be used to train and improve Facebook’s AI services. Meta, the company behind Facebook, offers a way for users to opt out of this data scraping, but it's not as straightforward as ticking a box.
Image Courtesy: Ideogram
How to opt out?
To opt out, start by visiting the help page provided by Meta. Look for the section that says, "I want to object to or restrict the processing of my personal information from third parties used for building and improving AI at Meta." You'll need to fill in a form with details such as your country of residence, name, surname, and email address. The form will ask if you believe Meta’s AI models have any knowledge of you. If they do, you’ll need to provide any prompts you entered that led to your personal information showing up in a response from a Meta AI model. Additionally, Meta asks for a screenshot to prove that your personal information appeared in a response from one of their AI models.
The last part of the form includes a section where you can give more context about your concern and specify what you are requesting. For example, you might explain why you don’t want your data used in this way.
Twitter user Tantacrul mentioned that Meta is also sending notifications to users about the new AI features. These notifications include a "right to object" link that directs users to a similar form. This form also asks how the processing of your data impacts you. Meta requires users to enter a One-Time Password (OTP) sent to their email address for security reasons. However, this step might discourage some people from completing the form, as the submission won’t be processed without the code.
Meta’s notice clarifies that if they honor your objection, it will only apply to future data processing. Tantacrul noted that Meta honored his objection, but many of his friends and users on Reddit encountered error messages when trying to submit the form.
Using user content to train AI models is becoming more common. For example, OpenAI recently partnered with Reddit to incorporate users' posts into ChatGPT.
In summary, if you want to opt out of Meta’s use of your public content for AI training, you need to fill out a detailed form on their help page. This includes providing specific information about how Meta’s AI has used your data and a screenshot for proof. Notifications about the new AI features also offer a way to object, but Meta's process might deter some users due to additional steps like entering an OTP. Despite some users experiencing issues, Meta has honored objections when properly submitted.