The US technology giant Meta has announced that it will not introduce multimodal AI models—so-called virtual assistants—in Europe due to regulatory uncertainties. A company spokesperson confirmed this decision to Euronews today.

We will release a multimodal LLaMA model in the coming months, but not in the EU due to the unpredictable nature of the European regulatory environment," said a Meta spokesperson.

This news, first reported by Axios, comes after Meta had already paused the introduction of its AI assistant in Europe. The Irish Data Protection Commission had instructed Meta to delay its plan to use data from adult Facebook and Instagram users to train large language models (LLMs).

Meta had updated its privacy policy, announcing that all public and non-public user data – with the exception of chats between individuals – would be used for current and future AI technology. This change was set to take effect on June 26.

As a result, the Austrian data protection organization NOYB filed complaints with the data protection authorities in eleven EU member states, claiming that Meta's practices are not compliant with the EU General Data Protection Regulation (GDPR).

NOYB requested an "urgency procedure" under EU data protection regulations and raised concerns about the processing of the personal data of approximately 4 billion Meta users.

Meta described the order as a "setback" for European innovations and emphasized that the company remains firmly convinced that its approach complies with European laws and regulations.

The Irish Data Protection Commission (DPC) stated to Euronews in June that "Meta has postponed the launch after a series of inquiries from the DPC, which have now been answered." Meta had granted users a four-week notice period before the original training, according to the DPC.

Meta has its own large language model called Llama, the latest version of which (Llama 3) was released in April and powers the Meta AI assistant.