YouTube will begin automatically labeling AI videos using internal detection signals, moving beyond reliance on creator disclosure

Starting this month, YouTube will use 'new internal signals' to flag videos showing 'significant photorealistic AI use,' automating a process that previously depended on creators to voluntarily indicate AI-generated content.

The system will identify AI content through C2PA metadata indicating a purely AI source and the use of watermarked Google tools like Veo. Labels applied for these two reasons are marked 'permanent' and cannot be appealed by creators.

YouTube introduced initial AI content labeling in 2024, but is now making labels more prominent as AI video models like Sora, Runway, and Google's Veo have improved in realism and consistency.