Meta Ai, what can do now those who have not opposed the use of their Instagram and Facebook data for training

The May 27th (today) Meta has made it back formally the training of generative artificial intelligence, Meta Aiwith the data and contents shared by European users on Instagram and Facebook. This means that all posts, Reels, comments and states of those who have not declared the opposite can be exploited by the company in order to train its models. A fundamental step so that they are able to learn the European « context » and therefore answer our questions in a relevant way. European regulation, As we explained here, He made to the destination to give the choice to each individual (European) user to be able to make their data for training available. Until yesterday, May 26, it was possible to fill in a form to be able to exercise your « right of opposition»And therefore prevent the contents that we have loaded or loaded in the future on our profiles enter the gigantic database for the training of Meta Ai. Now one opens one second phasewhere it is still possible to oppose, somehow, if you want, but with large limits.

What data uses destination to for training

In April Meta began to train destination ai With all the contents public – Shared by adult users – published on its social networks, Instagram and Facebook. At the same time, to repeat the European legislation, it made available a form through which it was possible to do « opt-out»From this operation, that is to oppose the use of your information for training. The form, if completed by May 26, was retroactive. And therefore it was valid throughout the amount of data that over the years we have posted, as well as for the future. Specifically, Meta Ai can exploit public data on Instagram and Facebookbut It cannot in any way access our conversations on WhatsApp and Messenger: these are protected by end-to-end encryption. Then there is another portion of data useful for artificial intelligence, or The interactions that users have a destination to Same chat or research on social platforms. In this case, Messenger is also included, while WhatsApp is not. To be specified that, at the moment, It is not possible to deactivate destination ai: you can only choose whether to use it or not.

What can do those who have not completed the form to oppose

The contents of those who have completed the form to oppose the use of their data for the training of Meta a by May 26, therefore, will be excluded from the database useful for the training itself. If, on the other hand, the form has not been completed, you can also do it later (Here the one for Facebook, Here the one for Instagram), but will no longer have retroactive value. It means that from that moment on our data will be excluded, but those loaded previously will instead be included. Then there is another form made available by Meta – This – «For third party information to use for the AI ​​di Meta». It is used to send requests regarding information that concern us but that have been shared by other users. They include, explain, « public domain data on the internet and information under license by someone who has the rights and authorizes destination to use them ». The form requires concrete examples of using this information from Meta Aiasks to provide prompts used and screenshots. This form can be useful if we find data concerning us in a chatbot response.