Meta to use Facebook, Instagram posts of EU users to train AI

Meta announced Monday that all public posts and comments by adult users on its platforms including Facebook and Instagram as well as their interactions with the company’s AI chatbot in European Union (EU) will now be used for training Meta AI. 

The social media giant has assured that public content of users below 18 years as well as private messages of all users will not be used for AI training. Also, adult users who don’t want their data to be fed into AI will be presented with an objection form. 

Meta claims that this will allow its AI models to have a better understanding of European cultures, languages and how they express themselves using humor and sarcasm. 

“It is so important for our generative AI models to be trained on a variety of data so they can understand the incredible and diverse nuances and complexities that make up European communities,” Meta said in a blog post. 

Though available in the US since 2023, Meta AI’s roll out in the EU was delayed due to regulatory and data privacy concerns. Meta had to pause the release of Meta AI in June 2024 after pushback from the Irish Data Protection Commission (IDPC).  

“Since then, we have engaged constructively with the IDPC and look forward to continuing to bring the full benefits of generative AI to people in Europe,” the company said in the blog post. 

EU’s General Data Protection Regulation (GDPR) is one of the strictest data privacy laws in the world with a robust enforcement mechanism that includes heavy fines. 

In November 2024, Meta was slapped with a fine of €798 million by the European Commission for engaging in anti competitive practices in its online classified ad business. 

GenAI based apps were also subject to more scrutiny in Europe than in other markets. For example, in March 2023, OpenAI’s ChatGPT was banned in Italy  by its data protection authority for illegally collecting user data and failing to provide  safe space for underage users. The ban was lifted in April after the regulator’s concerns were addressed by OpenAI. 

Europe is one of the key markets for US based social media and AI firms. According to estimates, Instagram had 269.1 million monthly users, while Facebook had 260.6 million users in Europe as of 2024. 

GenAI chatbots and apps are changing how people look up information online. It is helping them enhance their online conversations and social media profiles using AI-generated images and text. For instance, ChatGPT’s user base doubled within weeks of adding viral image generation tools that allowed users to transform their photos into Studio Ghibli-style artwork. 

All major AI firms including OpenAI, Google and Meta have been accused of training their models using illegally harvested data from social media, news articles, books, and research papers without seeking permission from users, publishers, or authors. 

According to court filings in an ongoing lawsuit against Meta, the company allegedly used LibGen, an online database of millions of pirated books, comics, magazines, and research papers, to train its AI models. 

Image credit: Flickr