Meta will start using data from EU users to train its AI models

Meta plans to start using their users from European Union users, to teach their company Announced today. The TECH giant, which began this week, will be able to report to Europeans and the family of these applications via e-mail. In addition, the notice will be closed to a form where a form users of a form to abandon the process. “We have made it easier to find, read and use this protest, and we will also honor all the protest, as well as the newly presented,” Meta.

The company notes that only social entries and meta will use the information collected by the AI ​​with the purposes of training. By the age of 18, users will not use any communication facilities, or any interaction messages prepared by users.

“We believe that it is not only available for Europeans. Therefore, it is set for them. Therefore, it is based on us to understand that the incredible and various nuances and inks formed for us and the European communities.

“This means everything from dialects and collocuents, hyper-local knowledge and humor and sarcasm in different countries. This is especially important, because it is more developed with multiple modal functionality covering text, sound, video and images.”

Meta notes include other AI companies, including Google and Openai, as well as using European users to correct their systems. Today’s announcement is in line with the initial release of the methane Llama 4 models. After some early hype was the company LMarena is charged in playing the gameA website where people compare the results of different AI models to sort it. When researchers saw, 4 of Meta Llama 4 have provided an experimental version to “optimized for negotiations”.