YouTube moves AI labels to prominent spots, auto-detects photorealistic AI

YouTube has announced a significant overhaul of its AI labeling system, moving disclosures from the hidden “How this content was made” section to prominent positions where viewers can see them instantly. For long-form videos, the AI label now appears directly below the video player, above the description, while for YouTube Shorts, it shows as an overlay. This change follows Google's expanded AI verification efforts at I/O and aims to provide users with immediate context about AI-generated or altered content. YouTube describes this as the single label format for all photorealistic and meaningfully AI-altered content; for unrealistic or slightly altered content, disclosures remain in the expanded description.

Beyond repositioning labels, YouTube is also rolling out new internal signals later this month to automatically identify and label photorealistic AI videos. While creators are still required to manually disclose AI use, the system will auto-apply labels if it detects significant photorealistic AI content and the creator hasn't disclosed. Creators can update incorrect labels via YouTube Studio, but labels become permanent if the content was created using YouTube's own AI tools like Veo or Dream Screen, or if it contains C2PA metadata indicating full AI generation. YouTube already uses markers like C2PA and Google's SynthID, but this represents a commitment to more proactive enforcement—at least for videos mimicking realistic humans and environments. Importantly, disclosure labels alone won't affect monetization or recommendations.