YouTube: AI labels to become more visible and automatic

YouTube is moving the labels for AI-generated content to more prominent positions. The video platform is also introducing an automatic detection system that can label videos even without the uploader's self-disclosure – for example, if the use of AI is evident from the C2PA metadata. YouTube announced this in a blog post.

Previously, the AI notice appeared below the video in the expanded description area. In the future, for normal horizontal videos, it will appear directly below the player, making it visible without further scrolling. For Shorts, the label will be displayed as an overlay directly in the video.

YouTube distinguishes between obviously AI-generated videos and those AI clips that could deceive viewers with their realistic appearance. The prominent placements of the AI notice will only be implemented for AI videos that appear photorealistic or have been significantly altered by AI, YouTube writes in the blog post. For animated AI videos, for instance, the labels will remain in the old position in the expanded description text.

Automatic AI Detection

Until now, YouTube has relied on a more or less voluntary self-disclosure by uploaders to label AI videos. Now, YouTube wants to detect AI clips independently. Creators can contest a wrongly applied label in YouTube Studio. However, if content was created with Google's own tools like Veo or contains AI markers in the metadata, such an AI label is considered irrefutable according to YouTube. C2PA stands for Coalition for Content Provenance and Authenticity and is an open standard that embeds information about content origin directly into the media file.

According to YouTube, the new system will be gradually implemented starting in May 2026. The existing disclosure obligation remains unaffected. Those who do not declare their use of AI should expect YouTube to set the label itself.

According to YouTube, an AI label has no influence on recommendation algorithms and advertising. The company justifies the changes with feedback from the community collected since the introduction of the disclosure obligation in 2024. Since spring 2024, creators on YouTube have had to disclose photorealistic AI content.

(dahe)