Goal He has begun to use the data of European users from their social networks, as announced last week. The accounts of the ‘Meta products’ (Facebook, Instagram, Threads, Messenger, WhatsApp) in Europe have started Receive notifications in which they are informed and given the option to exclude their data and public content of the target training AIthrough the corresponding app and also by email.
Mark Zuckerberg’s company, which is not the first time he uses the data of its European users to train their AI, It does not seem to have too much effort that users are aware. A notification is easily lost between the daily tens that generate their social networks for each user and something similar can be said of the emails, so it is expected that Many continue giving their content without being really aware of it or the scope. Goal uses all the public content of an account since it openedthat, in the case of its longest network, Facebookit covers up 2004. Twenty -one years.
The last time Meta stopped using them, at the request of the Data Protection Authority from Ireland, it was in May 2024. The company says it now has the approval of the European Data Protection Committee And that is why it is released again to the collection of data and content of the users to train their artificial intelligence models.
What data from IT uses goal to train your AI?
Goal explains in its privacy center that the public content of users includes:
- Name.
- Facebook and Instagram username.
- Profile photo.
- Activity in public groups, Facebook pages and channels.
- Activity in public domain content. For example, your comment, qualification or review in Marketplace or a public Instagram account.
- The content ‘that you can configure as “public”, such as publications, photos and videos that you publish in your profile or stories or reels’.
At this last point, The writing is ambiguoussince the same content usable for training also ‘can’ be configured as private. Strictly speaking, with that formula I could be using private accounts publications, something that does not expressly refer to at any time. Much clearer would have been to express “the content configured as a public” than the ‘that you can configure as “public”‘, which includes the one that can also configure as private.
In Facebook support It also appears in the public content categoryand therefore usable to train, the information provided when creating a profile. This is, in addition to the username and profile photo:
- Cover photo.
- The pronoun that Facebook uses when referring to you.
- Your networks (such as your school or workplace).
- The user identifier (account number) found in the URL of your profile.
- The age range.
- The language and the country.
What messes does specify that Does not include is the content of private messages with friends and family. Again, a striking clarification since Private messages can be with many other accounts as well as family and friends.
Cases to oppose the use of your data is useless
Goal ensures that it will comply with all the requests you receive, but There are some cases that are exceptions:
- When another user, who has not shown inconvenience in the use of its target content, publicly share an image in which you or your information appear.
- When a person mentions or mentions your information in publications or public texts.
- The content of your private messages can be public if any member of the conversation ‘decides to share them (your private messages) with our artificial intelligences’.
How to prevent the goal from using your data and contents to train goal AI
This is the target link provides in its emails and notifications to object to the use of your data by AI. You just have to enter the mail associated with your account and The objection will be effective for all those that appear added to the Meta Accounts Center. In the case of having other accounts that are not in that same account center, You must request the objection separately.
Meta also facilitates that you explain ‘How this data processing affects you’but it is optional.