Meta will withhold multimodal AI models from the EU amid regulatory uncertainty
The move could prevent anyone who planned to build services using Meta's AI models from offering them in Europe.
Meta has decided to not offer its upcoming multimodal AI model and future versions to customers in the European Union citing a lack of clarity from European regulators, according to a statement given by Meta to Axios. The models in question are designed to process not only text but also images and audio, and power AI capabilities in Meta platforms as well as the company’s Ray-Ban smart glasses.
"We will release a multimodal Llama model over the coming months, but not in the EU due to the unpredictable nature of the European regulatory environment," Meta said in a statement to Axios.
Meta’s move follows a similar decision by Apple, which recently announced it would not release its Apple Intelligence features in Europe due to regulatory concerns. Margrethe Vesteger, the EU’s competition commissioner, had slammed Apple’s move, saying that the company’s decision was a “stunning, open declaration that they know 100 percent that this is another way of disabling competition where they have a stronghold already.” Withholding Meta’s multimodal AI models from the EU could have far-reaching implications — it means that any companies that use them to build their products and services would be unable to offer them in Europe.
Thomas Regnier, an EU spokesperson, told Engadget that the regulator does not comment on individual decisions of companies. "It is the companies' responsibility to ensure that their services comply with our legislation," Regnier said in a statement and added that all companies are welcome to offer service in Europe as long as they comply with the bloc's laws, including the upcoming Artificial Intelligence Act.
Meta told Axios that it still plans to release Llama 3, the company’s upcoming text-only model in the EU. The company’s primary concern stems from the challenges of training AI models using data from European customers while complying with the General Data Protection Regulation (GDPR), the EU's existing data protection law. In May, Meta announced that it planned to use publicly available posts from Facebook and Instagram users to train future AI models but was forced to stop doing so in the EU after receiving pushback from data privacy regulators in the region. At the time, Meta defended its actions, saying that being able to train its models on the data of European users was necessary to reflect local culture and terminology.
"If we don’t train our models on the public content that Europeans share on our services and others, such as public posts or comments, then models and the AI features they power won’t accurately understand important regional languages, cultures or trending topics on social media," the company said in a blog post. "We believe that Europeans will be ill-served by AI models that are not informed by Europe’s rich cultural, social and historical contributions."
Despite its reservations about releasing its multimodal models in the EU, Meta still plans to launch them in the UK, which has similar data protection laws to the EU. The company argued that European regulators are taking longer to interpret existing laws compared to their counterparts in other regions.
Update, July 18 2024, 6:40 PM ET: This story has been updated to include a statement from an EU spokesperson.
