YouTube Introduces Tool for Labeling AI-Generated Content in Videos

YouTube has unveiled a new tool designed to boost transparency and build trust between creators and viewers on the platform. This tool will require creators to disclose when their videos contain content created or altered using artificial intelligence (AI) that could be misconstrued as real.

YouTube image

The new disclosure tool focuses on content that leverages AI to create realistic scenes, people, or places that viewers might mistake for genuine footage. This could include:

YouTube disclosure on AI-generated content

• Videos that digitally alter the likeness of a real person.
• Videos that manipulate real events or locations.
• Videos that use AI to create scenes for fictional purposes, but appear very realistic.

There are some exceptions to the disclosure requirement. Creators won't need to disclose the use of AI for productivity tools, such as automatic script generation or caption creation. Additionally, minor changes or clearly unrealistic content also fall outside the disclosure requirement.

The labels generated by the disclosure tool will be progressively rolled out across YouTube on various platforms and formats. Typically, these labels will appear in the description section below the video. However, for videos that address sensitive topics like health or news, the labels will be more prominent, appearing directly on the video player itself.

AI-generated content labels

The rollout of the disclosure tool will be gradual across YouTube. This timeframe is intended to provide creators with an adjustment period. However, YouTube has plans to enforce disclosure measures in the future. In cases where creators consistently fail to disclose the use of AI for realistic content, YouTube may take action, including adding a label themselves, especially if the content has the potential to mislead viewers.

YouTube is also working on an update to its privacy process. This update will allow individuals to request the removal of AI-generated content or other synthetic media that replicates an identifiable person, including their voice or face. More details about this specific process will be announced by YouTube soon.

YouTube is actively involved in industry-wide efforts to promote transparency when it comes to digital content. This includes their participation as a leading member of the Coalition for Content Provenance and Authenticity (C2PA), which reflects their dedication to fostering a more genuine and transparent online environment.

The disclosure tool will first be available on the YouTube mobile app, followed by desktop and TV versions.

In a statement announcing the update, YouTube emphasized the essential role creators play on the platform and their ongoing commitment to guide viewers through the evolving landscape of generative AI. YouTube acknowledges the dynamic nature of this field and their dedication to continuous learning and improvement. They believe that increased transparency will lead to a deeper understanding of how AI can be harnessed to empower human creativity.