Meta✴’s next multimodal AI model, as well as other future models, will not be available to customers in the European Union due to a lack of regulatory clarity, the company told Axios.
Image source: OpenClipart-Vectors/Pixabay
«We will release the Llama multimodal model in the coming months, but not in the EU due to the unpredictable nature of the European regulatory environment,” Meta✴ said in a statement sent to Axios. Meta✴ also said that European companies will not be able to use its multimodal models, even if they are released under an open license. In addition, companies from other countries will not be able to offer products and services in Europe that use the new multimodal models from Meta✴.
Apple followed suit, announcing last month that Apple Intelligence features would not be available in Europe due to Digital Services Act (DMA) requirements that could “reduce the security and privacy of user data.”
Meta✴ also plans to release a larger text version of the Llama 3 model soon, which will be available to companies in the EU.
According to Axios, Meta✴’s problem has more to do with how it may be using European customer data to train its model, thereby violating the EU’s General Data Protection Regulation (GDPR). The company has already been accused of illegal mass collection of personal data of European users.
In May, Meta✴ announced plans to use public posts from Facebook✴ and Instagram✴ users to train future models, sending more than 2 billion notifications to users in the EU. But then it was forced to abandon it due to a ban by the Irish Data Protection Commission (DPC) and the UK Information Commissioner’s Office (ICO).