Meta to train its AI models on public data in EU

Shortly after receiving approval from the European Union (EU), Meta is set to train its artificial intelligence (AI) models using public data from adult users on its apps, including Facebook, Instagram, WhatsApp, and Messenger.

This move comes following the release of  Meta AI's limited version in the EU in March, following its debut in the US and other global markets.

The Mark Zuckerberg-led company will begin collecting public posts, comments, and interactions with Meta AI for AI training purposes, excluding private messages and data from users under 18.

Users in the EU will receive in-app and email notifications in the upcoming days explaining how their data will be used, with an option to opt-out via a form.

It is worth mentioning that Meta has committed to honouring all objection forms, both new and previously submitted.

Moreover, the European Data Protection Board (EDPB) declared that Meta's approach meets legal obligations under EU data protection laws.

Notably, Meta's decision to train its AI models on public data follows similar practices by Google and OpenAI, which have also used European user data to train their models.

The company has also emphasised the importance of diverse training data to build AI that understands European cultures, languages, and nuances.

Notably, Meta had paused AI training last year due to regulatory pressure but has now engaged constructively with the Irish Data Protection Commission (IDPC) to move forward with its plans.