ButSpeak.com
News which Matters.
In a significant development, Meta Platforms, the parent company of Facebook, announced on Monday its intention to incorporate publicly shared social media content from Europe into the training of its generative AI models. This move marks a shift towards aligning Meta’s data usage practices in Europe with those in other parts of the world, despite the stringent privacy regulations in the EU.
Meta plans to use content from public posts on platforms like Instagram and Facebook to train its Llama large language models, as detailed in a company blog post. This approach comes after a period of cautious deliberation due to the EU’s rigorous privacy and transparency requirements.
In a September interview with Reuters, Meta’s top policy executive explained that while the company already uses public posts from Facebook and Instagram globally to train its AI models, it has been careful to exclude private posts and messages. As of April, during the release of the latest Llama versions, Meta was still determining the best way to implement this strategy in Europe.
To address transparency concerns, Meta announced last month that it would begin notifying users in the European Union and the United Kingdom about how their public information on Meta services is utilized for AI development and improvement.
Despite these notifications, advocacy group NYOB (None of Your Business) has raised objections, filing complaints across various European countries. They argue that the notifications fall short of EU privacy regulations, which require Meta to obtain explicit opt-in consent from users before using their data.
This decision by Meta underscores the ongoing challenges and debates surrounding data privacy and AI development, particularly within regions with stringent regulatory frameworks. As Meta navigates these complexities, its approach to incorporating European social media content for AI training will likely continue to evolve in response to both technological advancements and regulatory demands.
