Meta will not launch its most advanced AI in the EU due to the strong regulatory environment

In addition to being a multimodal AI, Llama 3 is expected to be present in different products, such as Ray-Ban Meta smart glasses, which would allow its integration into more everyday uses and would resemble the real-time assistants seen in science fiction films.

In fact, it emerged Thursday that Meta is in talks to acquire a stake of about 5% in EssilorLuxottica, its production partner for smart glasses, The Wall Street Journal reported Thursday, citing people familiar with the matter.

However, according to the AI ​​Law passed in March, certain uses of AI are prohibited, such as those that “threaten the rights of citizens.” This category includes biometric categorization to deduce information, random identification of faces to build databases, and even systems that try to read emotions in jobs or schools.

The regulation also establishes strong transparency requirements for developers, and companies will have until August 2026 to implement the corresponding changes to their technologies to comply with the Law.

Tech companies can be fined for breaching the rules, with the highest being around $38 million or 7% of global annual revenue.

It is for these reasons that Meta has decided to exclude the EU from the launch of its next model, something that Apple has also considered, as it said its Apple Intelligence system would probably not be available in this territory due to concerns about the Digital Markets Act.

It should be noted that for years, the EU has been waging a tough battle against Meta over the way it collects and uses user information, mainly for its targeted advertising business, and in that territory it implemented a paywall to prevent ads on its social networks.

In the context of AI, the company was called out last month for using its users’ data to train its large language models, something that people did not like, who mobilized to revoke their consent to allow the company to use their data for these purposes.

Meta said that this type of information is necessary for the development of technology, mainly to adapt it to the precisions of language and cultural differences or current issues. However, people are increasingly aware of the value of their data and have become opposed to this type of measures.