Meta has decided not to offer its upcoming multimodal AI model and future versions to customers in the European Union due to the lack of clarity from European regulators, it said in a statement filed by Meta to the European Commission. AxiosThe model in question is designed to process not only text but also images and audio, and powers the AI capabilities of the Meta platform and the company's Ray-Ban smart glasses.
“We plan to release the multimodal Llama model in the coming months, but it will not be released in the EU due to the unpredictable regulatory environment there,” Mehta said in a statement. Axios.
Meta's move follows a similar decision by Apple, which recently announced it would not release its Apple Intelligence feature in Europe due to regulatory concerns. EU Competition Commissioner Margrethe Vesteger slammed Apple's move, calling it “an astonishing, public declaration that they 100% understand that this is another way to kill competition where they are already based.” Withholding Meta's multimodal AI models from the EU could have far-reaching implications: companies that use the models to build products or services would no longer be able to offer them in Europe.
EU spokesman Thomas Regnier told Engadget that the regulator doesn't comment on individual companies' decisions. “It is up to companies to ensure that their services comply with EU law,” Regnier said in a statement, adding that all companies are welcome to offer their services in Europe as long as they comply with EU law, including the upcoming artificial intelligence law.
Mehta said Axios The company said it remains committed to releasing its upcoming text-only model, Llama 3, in the EU. The company's main concerns stem from the challenge of training its AI models with data from European customers while complying with the EU's existing data protection law, the General Data Protection Regulation (GDPR). Meta announced in May that it planned to use public posts from Facebook and Instagram users to train future AI models, but faced backlash from EU data privacy regulators, forcing it to halt its EU rollout. At the time, Meta defended its actions, saying being able to train models with European users' data was necessary to reflect local culture and terminology.
“Without training our models on public content, such as public posts and comments that Europeans share on our and other services, our models and the AI features they power will not accurately understand important regional language, culture, and trending topics on social media,” the company said in a blog post. “Europeans would be disadvantaged by AI models that do not take into account Europe's rich cultural, social, and historical contributions.”
Meta has been cautious about releasing its multimodal model in the EU, but plans to do so in the UK, which has similar data protection laws to the EU. The company argued that European regulators have been slower to interpret existing laws than their counterparts in other regions.
Update, July 18, 2024 at 6:40 PM ET: This story has been updated to add a statement from an EU spokesperson.
