Goal will lead to its generative artificial models (AI) (AI) about the public content of its European users and their conversations with the goal AI Chatbot, the group announced in a statement on Monday. In particular, the company will operate publications and public comments shared on its social networks to train its models.
For the company, it is not only necessary to provide an AI to Europeans, but also ensure that they can understand their dialects, family expressions or local knowledge, which cannot be done without training their models in their public data.
Possible opposition
As of this Monday, April 14, Meta will begin informing users about the use of their public content to train their models through notifications. These will explain the type of data used for them, but also how this training will improve their AI and the experience of the global user.
However, European users will have the possibility of opposing the use of their data for this purpose, thanks to a link that refers to a dedicated form in these notifications.
The company also specifies that public accounts data in the EU will not be used to train its AI models. This announcement occurs almost a month after the launch of goal AI in the EU. While the Californian group launched its chatbot in 2023 in the United States, it took it to make it available to the EU due to regulations.
“Last year, we delayed the training of our main language models using public content, while regulators clarified the legal requirements. We greet the opinion presented by the European Data Protection Committee in December, which confirmed that our initial approach was in accordance with our legal obligations,” underlined, rejoiced to “continue with the Europeans to benefit from all the advantages of the generative year.”
Source: BFM TV
