Meta Refuses to Launch Multimodal AI Models in the European Union

Meta, the parent company of Facebook and Instagram, has decided not to launch its new multimodal AI models in the European Union. This decision is due to ambiguity in the EU's laws and regulations regarding artificial intelligence. These models are capable of understanding and generating video, audio, images, and text, and were intended to be used in various Meta products, including smartphones and smart glasses. This move by Meta means that European companies will also be unable to use these models, even if they are released under an open license. However, Meta plans to soon release a larger, text-only version of its Llama 3 model, which will be available to customers and companies in the EU. The issue for Meta is not with the AI law, which is still being finalized, but with how to train the models using data from European customers while complying with the GDPR, the EU's data protection law. In May, Meta announced its intention to use public posts from Facebook and Instagram users to train future models, but in June, after announcing its plans publicly, Meta was ordered to stop training data in the EU. The UK has a law almost identical to the GDPR, but Meta states it does not face the same level of regulatory uncertainty and plans to launch its new model for British users. A Meta representative told Axios that European regulators spend much more time interpreting existing laws compared to their counterparts in other regions. This move by Meta increases tensions between American tech giants and European regulators, which have long existed over privacy and antitrust regulations. Tech companies argue that these regulations harm consumers and the competitiveness of European firms. However, a Meta representative told Axios that training data in Europe is key to ensuring that the company’s products accurately reflect the region's terminology and culture.