YouTube has announced it will now automatically label videos that use significant photorealistic AI, according to the company. The change comes as part of its ongoing efforts to manage AI content on its platform, following the release of Google’s Gemini Omni AI models at Google I/O. Starting in May, YouTube will use new internal signals to identify AI-generated content and apply labels accordingly. Creators are still encouraged to disclose their use of AI, but if they fail to do so, YouTube will label the video for them.
The move follows the expansion of YouTube’s AI deepfake detection, which now allows any adult to scan YouTube for face matches after initial tests with celebrities, public figures, and politicians. YouTube says it will also make its AI labels more consistent and prominent. Previously, labels appeared in the expanded description unless the video touched on sensitive topics like health or news, in which case a prominent label would appear directly on the video.
Now, labels will appear directly below the video player above the description for long-form videos and overlay directly on YouTube Shorts. Labels will also be permanently attached to videos with C2PA metadata indicating full AI generation. OpenAI recently joined the C2,PA standard, joining Nvidia, Kakao, and Eleven Labs.
The company said moving the labels will make them more obvious to users encountering photorealistic, AI-altered, or AI-generated content. AI labels won’t affect video recommendations or monetization. In addition to policing AI content, YouTube has been investing in AI for features like interactive search, playlist generators, and AI video summaries.
Source: techcrunch
