Meta to Use Public Facebook and Instagram Data for AI Training in the UK

Meta, the parent company of Facebook, has announced plans to utilize public photos and texts shared by adult users on Facebook and Instagram in the UK to train AI models in the coming months.

This move aims to ensure that the generative AI model reflects British history, culture, and idioms, and that British companies and institutions can leverage the latest technology. Meta made this announcement on its official website, explaining the reasoning behind this decision.

Meta has been working closely with the UK Information Commissioner’s Office (ICO) on this matter, following the regulator’s request to pause AI model training in the UK. The company welcomes the ICO’s constructive approach to AI model training.

Meta has assured users that it will not use personal messages shared with friends or family to train its AI, nor will it use information from accounts of users under the age of 18. Instead, the company will rely on publicly available information, such as public posts, comments, public photos, and captions, to improve generative AI models for users in the UK and worldwide.

Starting next week, adult users of Facebook and Instagram in the UK will begin receiving in-app notifications that will also allow them to opt out of having their data used to train generative AI models.

In May, Meta announced its intention to use data from its social media users to train AI. The company updated its privacy policy to use data collected from Facebook, Instagram, and other public posts and captions for AI training.

However, this plan has faced resistance from regulators globally. European regulators have asked Meta to delay training its large-scale language model (LLM) using open-source content, leading to the temporary suspension of the launch of Meta AI in Europe.

Meta temporarily suspended its generative AI service in Brazil after authorities banned the use of user data to train AI, but later decided to comply with the authorities’ strict requirements.

A recent controversy in Australia revealed that Meta had been training its AI models using users’ content without permission, and it is expected that the Australian government will move to revise related laws.