With the European Union’s AI Act set to require AI content labeling and machine-readable provenance markers beginning August 2026, YouTube has unveiled a series of changes aimed at giving viewers clearer information about synthetic content on its platform. The company announced on May 27 that it will now automatically apply disclosure labels when its internal systems detect photorealistic AI-generated content that creators have not voluntarily flagged. Until now, the platform relied primarily on self-reporting by creators to identify such material.
Labels will also become more prominent. For long-form videos, the disclosure will appear directly below the player, and for Shorts, it will be displayed as an overlay. Previously, these labels were largely confined to expanded video descriptions, making them easy to overlook. Some labels cannot be removed. Videos produced with YouTube’s own generative tools, Veo and Dream Screen, will always carry a disclosure, as will content containing C2PA metadata that confirms full AI generation. In these cases, even a creator’s appeal through YouTube Studio will not result in the label being removed.
The platform stressed that AI-labeled content will not be disadvantaged in recommendation systems or lose eligibility for monetization, framing the changes as a transparency initiative rather than a restriction on creators.
YouTube has also widened access to its deepfake reporting tools, which were previously limited to public figures. All adult users can now flag face-based synthetic content for review, with voice cloning detection capabilities planned for later in 2026. The company said the updates reflect its goal of providing context about AI involvement in content creation without limiting how such content is distributed or monetized on the platform.