Meta, the parent company of Facebook and Instagram, has announced that it will suspend its plans to train its artificial intelligence models with the content uploaded by European users on its platforms. This decision was made following an intense exchange between Meta and the Irish data protection authority (DPC), where Meta has its European headquarters. The DPC welcomed the move, stating that it had sparked opposition from citizens and civil society groups who were concerned about their privacy rights being violated.
One of the main reasons for this opposition was that Meta did not offer users the option to consent to the use of their content for training AI models. Instead, it assumed this consent by default. This complicated targeting mechanisms further frustrated users’ ability to opt out of this practice.
Groups like Citizen 8 in Spain launched campaigns to raise awareness and provide simpler alternatives for users to prevent Meta from using their data. One criticism was that even if users objected to their image being used, there was no guarantee that it would not end up being used for training AI models if someone in their contact uploaded a photo or video with them in it.
Meta expressed disappointment at the DPC’s decision but emphasized that it complies with European data protection regulations and laws. The company defended its use of public content to build language models, stating that it does not use private messages or content from minors. However, its Privacy Director criticized the decision as a step backwards in European innovation and competitiveness in AI development.
The suspension of these plans comes after a series of legal challenges brought against Meta by groups such as Noyb who sought precautionary measures regarding the use of personal data without explicit consent.
