by Meta has announced that it will withhold its upcoming multimodal AI model, and future ones, from customers in the European Union due to unclear regulatory guidelines, Axios reports.
Why It Matters: This decision sets up a potential confrontation between Meta and EU regulators, showcasing a growing trend among U.S. tech companies to limit product availability in Europe over regulatory concerns.
Current Situation: Meta stated, “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.” This follows Apple’s recent decision to withhold its Apple Intelligence features from Europe for similar reasons. The Irish Data Protection Commission, Meta’s primary privacy regulator in Europe, has not commented on the issue.
Driving the News: Meta plans to use the new multimodal models, which can interpret video, audio, images, and text, in various products, including smartphones and Meta Ray-Ban smart glasses. However, European companies and those outside the EU offering products and services in Europe will not be able to use these models, despite them being released under an open license.
Future Availability: Meta will still release a larger, text-only version of its Llama 3 model for EU customers and companies.
Regulatory Challenges: Meta’s concerns are less about the upcoming AI Act and more about compliance with the General Data Protection Regulation (GDPR). In May, Meta announced plans to use publicly available posts from Facebook and Instagram to train future models, informing over 2 billion EU users about opting out. Despite briefing EU regulators in advance and addressing their minimal feedback, Meta was ordered to pause training on EU data in June, followed by numerous questions from regional data privacy regulators.
Comparative Landscape: The U.K., with a similar law to GDPR, has not posed the same regulatory uncertainties, allowing Meta to proceed with its model launch there.
Big Picture: Meta’s decision highlights the ongoing tension between U.S. tech giants and European regulatory bodies. The EU’s stringent privacy and antitrust regulations are seen as a hindrance by tech companies, who argue that these rules negatively impact consumers and the competitiveness of European firms.
Meta’s Position: A Meta representative emphasized the importance of training on European data to ensure products reflect local terminology and culture. Meta also pointed out that competitors like Google and OpenAI are already training on European data.
This development underscores the growing friction between innovation and regulation, as tech companies navigate complex legal landscapes to deliver their products globally.
