Meta plans to withhold its future multimodal AI models in the European Union due to lack of clarity from regulators, according to an Axios report. Multimodal models can process images and audio and power AI applications on Meta as well as on Ray Ban smart glasses. 

The company confirmed their decision in a statement to the outlet saying, “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.” 

The EU is particularly concerned about how Meta and other AI companies will train their AI models using data from customers while complying with their data protection law, the General Data Protection Regulation (GDPR).

Meta also shared that they still plan to release their next text-only model, Llama 3 in the EU.

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

Earlier in May, Meta said that it plans to train its AI models on publicly available posts of European users on Facebook and Instagram. But the company had to quickly halt this after criticism from data privacy regulators in the region. 

Apple has made a similar choice after it recently announced that will avoid releasing Apple Intelligence in Europe due to regulatory concerns.