YouTube Begins Automated Prominent Labeling of AI Videos

YouTube will begin using more prominent labeling for AI videos and will no longer rely entirely on uploaders to disclose when they use AI tools. The platform debuted its initial AI content labeling in 2024, which proved largely ineffective as AI videos at the time often revealed themselves through bizarre or disjointed appearances.

Advancements in AI models have since improved realism and consistency, prompting the update. Starting this month, YouTube will employ new internal signals to automatically flag videos showing significant photorealistic AI use. Creators must still indicate AI tool usage when uploading, but the new system adds an automated layer to the process.

Google remains vague about the specific internal signals involved in its AI detection. The blog post identifies two ironclad triggers that will result in permanent labels: C2PA metadata indicating a purely AI source and the use of watermarked Google tools like Veo. Videos flagged under these conditions cannot have their labels appealed.

For other cases where a video is incorrectly tagged as AI-generated, creators can submit an appeal. The updated labels address a key shortcoming of the 2024 system, in which voluntary disclosures appeared only in the expanded video description under a section titled “How this content was made.” Most viewers never encountered them.

The new labels take a far more visible position. On standard landscape videos, the AI tag appears directly below the video player and above the description box. For YouTube Shorts, it displays as a small overlay at the bottom of the video itself, adding to the existing visual clutter.

The label design consists of a small ellipse featuring “AI” text alongside an information symbol. YouTube intends for it to be clear and glanceable. Three distinct styles of these labels will appear across the platform.

Importantly, the automated prominent labeling applies only to photorealistic and meaningfully AI altered or generated content. Animated videos created with AI, unrealistic clips, or realistic videos containing only a few AI elements will continue to carry disclosures solely in the expanded description box rather than receiving the new visible tags.