BRUSSELS - American tech giant Meta said on Monday that it will begin using publicly shared content from users in the European Union (EU) to train its artificial intelligence (AI) models, Xinhua reported.
The company said in a statement that it will use public posts, comments, and interactions with Meta's AI features from adult users across its platforms, including Facebook and Instagram.
The move follows the launch of Meta AI in the region last month.
The company argued that training its models on European data is necessary to reflect the region's languages, humour, and cultural context.
EU users will begin receiving in-app notifications and emails this week regarding the data training policy, Meta said, adding that users can also opt out of the plan through a form it described as simple and accessible.
The company stressed its approach is consistent with industry practices, citing Google and OpenAI and emphasised that its use of data is more transparent.
Meta AI, as a standalone app, was launched in the United States in early 2024, but its expansion into Europe was delayed amid regulatory concerns.
A planned roll-out in 2024 was postponed due to the EU's stringent data protection and transparency requirements. - BERNAMA