It should be easier to tell at a glance whether a YouTube video contains AI-generated filth.
YouTube is looking to make it easier for users to know if a video was made using generative AI tools. The platform already requires creators to disclose any use of realistic-looking AI. Now, YouTube will scan videos for signs of AI-generated content.
If a creator has not disclosed whether they used genAI tools and YouTube’s systems “detect significant photorealistic use of AI,” the platform said it will automatically apply an AI label to their video. If the creator believes the label was included in their video in error, they can update their statement. However, if YouTube detects that a video was made using Google AI tools such as Dream Screen or Veo, or that it contains C2PA watermarks (an industry standard used to flag genAI creations), the label will remain in place permanently.
YouTube also suggests that it will be easier for viewers to see this AI label, which denotes the use of “photorealistic and significantly edited or AI-generated content.” The platform makes the label more visible in videos, placing it right below the video player, and on Shorts, where it will appear as an overlay.
More transparency (which YouTube says users are asking for) is welcome here, especially for those who prefer to avoid generative AI as much as possible. It would be nice if YouTube also started placing these labels on thumbnails in search results and suggestions to make it even easier to avoid such filth.
