Meta confirmed on Friday that it will begin training its artificial intelligence (AI) models on the public data on Instagram and Facebook in the UK. The company first planned to use this data three months ago, however, the plan was halted due to regulatory concerns. However, the social media giant confirmed that it has now received the nod to proceed with the AI training. The company claimed that this would enable the AI models to reflect the cultures of British users more accurately.
Meta to Train AI on UK Users' Social Media Data
In a newsroom post, Meta highlighted that it had previously paused the plans to train AI models with user data in the UK after regulatory authorities of the country raised concerns over how the data would be used once collected and how users could opt out of it. However, it claimed that the company had “engaged positively” with the Information Commissioner's Office (ICO) and was restarting the plan. Meta stated that it will begin training AI with this data over the coming months.
The social media giant claimed that the reason behind training Meta AI on this data was to reflect British culture, history, and idiom in a better way when people interact with the AI. Further, it added that the tech giant was aiming to reflect the diverse communities around the world by training on localised user data.
Highlighting how the company will go about collecting data, Meta clarified that it will not collect information from users in the UK who are under the age of 18. Private data such as messages sent to friends and family will also not be used to train AI. “We'll use public information – such as public posts and comments, or public photos and captions – from accounts of adult users on Instagram and Facebook,” it added.
Starting next week, adult users based in the UK will begin receiving notifications on Facebook and Instagram which will explain the details of what Meta is trying to do and how. Additionally, the notifications will also explain how users can access an objection form to prevent Meta from using their data to train generative AI models. Those who have already objected to this will not get any notifications and their data will not be used.
