YouTube is making it clearer when videos use AI. The platform will now automatically add labels to content with significant photorealistic AI, instead of relying only on creators to disclose it.
Labels are easier to spot. For long videos, they show up below the player. For Shorts, they appear as overlays. This way, viewers can see if AI was involved.
Creators are still expected to disclose AI use. If they don’t, YouTube will step in. Misidentified creators can update their disclosure, though labels stay if the video was made with YouTube’s own AI tools like Veo or Dream Screen.
Videos with C2PA metadata, which confirms full AI generation, will carry permanent labels. Photorealistic or altered AI content gets visible tags, while unrealistic or animated AI (like fantasy creatures) will only be labeled in the expanded description.
YouTube says these labels won’t affect monetization or recommendations. The goal is transparency, not penalties. At the same time, the company expanded its deepfake detection, letting any adult user scan for face matches, not just celebrities.
Also Read: YouTube rolls out option to disable Shorts on mobile
Beyond labeling, YouTube is also building AI features such as Ask YouTube, playlist generators, video summaries, and generative creation tools.
