As AI-generated media becomes more photorealistic, it’s increasingly important that people have tools to help them identify it. We’ve been working to put this principle into practice at Meta since April 2024, when we announced our approach to identifying and labeling AI-generated content on our platforms. More recently, we launched a research demo of a detection tool to help people better understand if an image was made with Meta AI as part of these efforts.
Signing the EU AI Act Code of Practice on Transparency of AI-Generated Content, which we confirmed we will do earlier this week, reflects our commitment to this principle. It also aligns with our ongoing work with peers across the entire stack to tackle these cross-industry challenges. Through forums like Partnership on AI and the Coalition for Content Provenance and Authenticity (C2PA), we’re working to develop solutions for identifying AI-generated content that are durable, technically-feasible and sustainable.
Getting this right is critically important. It means transparency measures that provide clarity, rather than more confusion. We must avoid a situation where online content is subject to a growing array of different labels and disclosures that end up overwhelming people, while at the same time adding additional regulatory complexity for providers.
As technical standards continue to evolve, we will work with the AI Office and other partners to keep the approach practical, interoperable and genuinely useful for the people it’s designed to serve.
