YouTube to Automatically Detect and Label AI-Generated Videos

YouTube is implementing an automated system to detect and label videos containing AI-generated content, aiming to provide greater transparency to its users. The video platform, which already requires creators to disclose their use of generative AI tools, will now analyze videos for signs of AI-generated material. When YouTube's detection systems identify significant photorealistic AI use in a video where the creator has not disclosed such use, the platform will automatically apply an AI label to the video.

The labeling system distinguishes between different types of AI detection. If a creator disputes an automatically applied label, they can update their disclosure to challenge it. However, labels will remain permanently on videos that YouTube detects were created using Google's own AI tools, such as Dream Screen or Veo, or that contain C2PA watermarks—an industry standard used to identify generative AI creations.

YouTube is also enhancing the visibility of these AI labels to help users quickly identify AI-generated content. The platform will place labels directly underneath the video player for standard videos and display them as overlays on YouTube Shorts. These labels specifically denote videos containing "photorealistic and meaningfully AI altered or generated content." The move addresses user demand for greater transparency regarding artificial intelligence in video content, though some observers note that expanded labeling on search results and suggested videos could further improve discoverability and help users avoid AI-generated content more easily.