Particle.news
Download on the App Store

YouTube Will Automatically Label Videos It Detects as Photorealistic AI

The move shifts labels from hidden descriptions to visible tags below the player or as overlays on Shorts, tying some labels to provable provenance metadata.

Overview

  • YouTube announced on Wednesday that it is rolling out internal detection signals to automatically apply AI disclosure labels when it detects significant photorealistic or meaningfully AI‑altered video.
  • The platform is moving tags out of the expanded description to a prominent position under long‑form players and as an on‑screen overlay on Shorts so viewers see AI context at a glance.
  • Labels that rely on machine‑readable provenance will be permanent, including videos with C2PA metadata or those created with YouTube’s own tools such as Veo and Dream Screen, while creators can dispute most other automated tags in YouTube Studio.
  • Reports note that Google described two clear triggers—C2PA metadata and watermarked Google tools—but left other internal detection signals vague, raising concerns about false positives and how large‑scale enforcement will work.
  • YouTube says labels will not directly change recommendations or monetization, though viewer reactions to visible AI tags could still alter a video’s performance, and the update responds to growing use of advanced generative video models and tightening provenance rules under regulators.