Meta to endorse EU code on transparency of AI-created content

With the rise of photorealistic AI-generated media, equipping people with the means to identify such content has become increasingly crucial. At Meta, efforts to implement this practice began in April 2024 when the company unveiled its strategy for marking AI-created content on its platforms. Recently, Meta introduced a research demonstration of a tool designed to help users determine if an image was produced using Meta AI as part of these initiatives.

Earlier this week, Meta confirmed its intention to sign the EU AI Act Code of Practice on Transparency of AI-Generated Content. This decision highlights the company’s dedication to transparency principles and aligns with its collaborative efforts across the industry to address these widespread challenges. Through engagement in groups like the Partnership on AI and the Coalition for Content Provenance and Authenticity (C2PA), Meta is working on developing lasting, technically viable, and sustainable solutions for identifying AI-generated content.

Ensuring these measures are effective is critically important. The goal is to provide clarity through transparency measures, rather than adding to confusion. It’s essential to prevent a scenario where the internet is flooded with a multitude of labels and disclosures that could overwhelm users and introduce more regulatory complexities for content providers.

As technical standards continue to develop, Meta plans to collaborate with the AI Office and other partners to maintain an approach that is practical, interoperable, and genuinely beneficial for the intended audience.