
German court allowed Meta to use user data for AI training
Germany’s judicial system made an important decision that could become a precedent for all of Europe. Now company Meta can use publicly available information of Facebook and Instagram users for training its artificial intelligence systems. Meta plans to use names, profile photos, comments and likes of adult users for training language models Llama. And chatbot Meta AI integrated into WhatsApp. Private messages and chats will not fall into training data. Yes, we believe them at their word. Especially after the same Meta is now being sued in the USA because of using pirate sites. And also conscious lying under oath and deliberate concealment of such facts.
The court rejected an urgent lawsuit from a consumer rights protection group. And representative of the consumer group Kristin Steffen emphasizes information already used for training is practically impossible to remove from the system.
The court ruled that even when removing identifiers, data remains personal. But Meta’s interests in developing artificial intelligence are recognized as more significant than privacy concerns.