Meta to Use ⁣Public Content for AI Training; EU ⁢Users ⁣Can Opt Out

BARCELONA – ‍Meta plans to leverage user-generated content from its platforms, including Instagram and Facebook, to train its artificial intelligence (AI) models, such as Meta AI, ⁢the WhatsApp assistant, ⁣and features within Instagram and Facebook Messenger.The company announced these plans in a statement⁢ released April 14,stating it will use ⁣public posts ⁤and comments shared by adult users within⁢ the European Union.

According to Meta, interactions with Meta AI, including questions and prompts, will also be used for training purposes. This initiative will utilize publicly available data ⁤and accounts of users over 18 to enhance the capabilities of ⁤MetaS AI.

Though, ⁢users have the option to object to the use of their data. Meta asserts that its data processing adheres to ⁣European regulations. ⁣A similar project was paused in June 2024 due to concerns raised ‍by EU authorities. Meta postponed the launch of its AI ⁤models in Europe following a request from the Ireland Data Protection Commission,its primary regulator in the region.

Opting Out of Data Usage on ‍Instagram

users can object to the use of their public data by submitting a request through this link.Upon submission,users will receive ⁣an email confirming their decision. Alternatively, users can navigate directly through Instagram by going to “Settings > Privacy Center” and following the provided link to opt out.

Opting Out of Data ⁤Usage on Facebook

Similarly, facebook users can object⁢ via this link or directly through Facebook by going to “Settings > Privacy ⁤and Security > Your Privacy > Privacy Center” and selecting⁤ the opt-out link. In both instances,only⁤ an email address is required,with an optional field for ⁣providing reasons for the decision.

Meta has stated that private messages between users and their contacts will not be used to train its generative AI models. Furthermore, ‍public data‍ from EU accounts belonging to individuals⁣ under 18 will also be excluded from training datasets. The company maintains that its approach is more clear than others in the ⁢field.