Meta has officially announced its signing of the Code of Practice under the EU AI Act regarding the transparency of AI-generated content. This decision marks a formal commitment by the company to implement tools that help users identify photorealistic synthetic materials on its platforms.
Work on implementing this principle has been underway since April 2024, when the company first announced its approach to labeling such content. Recently, as part of these efforts, a research demonstration tool for detection was launched, aimed at clarifying whether an image was created using Meta AI technologies.
The signing of the code occurs against a backdrop of increasing realism in generated media, making the presence of verification mechanisms critically important for users. The company's actions are directed toward the practical realization of regulatory requirements to ensure clarity regarding the origin of digital assets.
editorial commentary
Why it matters
The likely consequence will be an accelerated implementation of hidden labeling across all of the company's products for the global market to avoid fragmentation of standards. The next observable signal will be updates to the platforms' terms of use and the appearance of explicit visual indicators for users. The main uncertainty remains the effectiveness of these measures against malicious removal of metadata by third parties.