YouTube will automatically identify videos published with realistic content generated or modified by artificial intelligence (AI) and will add an identifying label if the content creator has not done so themselves; furthermore, it will display these labels in more prominent locations.
The video streaming platform has offered an AI labeling option since 2024, allowing content creators to indicate whether realistic videos shared with their audience were created or altered using generative AI. This initiative aims to ensure “greater transparency” regarding the content being viewed and to prevent confusion.
However, YouTube now intends to go a step further, announcing that it will begin automatically detecting AI-generated content by implementing new internal signals. Consequently, if AI-generated content is detected—and the creator has failed to specify their use of this technology—the company will automatically apply an indicative label.
The company shared this update in a blog post, detailing that it has gained insights into “what people find useful” regarding AI disclosures. Based on these findings, the platform aims to facilitate a process that is “much simpler and more intuitive for creators and viewers.”
Specifically, while YouTube still requires creators publishing videos to manually indicate when they have used realistic AI, starting this May, if the platform’s systems detect a “significant” use of photorealistic AI, they will automatically apply an AI label.
The platform clarified that if creators believe their content was misidentified as AI-generated, they can update the disclosure status within YouTube Studio. However, it also noted that automatic AI labels will remain permanent in certain instances.
For example, YouTube specified that these labels will always be displayed whenever the content was created using YouTube’s own AI tools, such as Veo or Dream Screen. Likewise, AI labels will also remain in place whenever content is identified as containing C2PA metadata.
AI LABELS IN MORE PROMINENT LOCATIONS
In addition to the above, YouTube has also announced that it will begin displaying disclosure labels for photorealistic content—as well as content that has been altered or generated using AI—”in a significantly more prominent location.”
For long-form videos, the company has specified that these labels will now appear “directly below the video player” and above the description. For short-form videos, the label will appear as an overlay directly on the video itself.
This approach allows viewers to identify the type of content at a single glance, further clarifying when content has been generated by AI and helping to prevent potential deception or misunderstandings.
“This is now the standard label format for all photorealistic content—as well as content modified or generated using artificial intelligence—on YouTube,” the company stated. However, for content that is not photorealistic, is animated, or has been only slightly modified, users will be able to find this information within the video’s expanded description.
Finally, it is worth noting that the presence of an AI disclosure label on its own “does not alter how a video is recommended, nor does it affect its ability to generate revenue,” as the company has clarified.
