YouTube Automatically Labels AI Videos to Fight Deepfakes

YouTube is taking a much stronger stance on AI-generated content as the platform faces a growing wave of realistic synthetic videos, deepfakes, and misleading media. In a major update announced this week, the company revealed that it will now automatically label videos that make “significant photorealistic AI” use, even if creators fail to disclose it themselves. The move marks one of the most aggressive transparency efforts yet by a major social media platform as generative AI tools rapidly improve and become more accessible.

For the past two years, YouTube has required creators to disclose when their content used realistic AI that could potentially confuse viewers into believing a person, place, or event was real. These labels were originally introduced through Creator Studio tools and generally appeared in the expanded video description unless the topic involved sensitive categories like news or health. However, YouTube now says that system is no longer enough in a world where AI-generated videos are becoming increasingly sophisticated.

The company explained that viewers have consistently asked for greater transparency around AI-generated media. To address those concerns, YouTube is introducing new internal detection systems capable of identifying AI-generated or AI-altered content. If a creator does not voluntarily disclose their use of AI, YouTube’s automated systems will apply the label on their behalf. According to the company, the policy itself has not changed, but enforcement is becoming more proactive as the amount of AI content on the platform explodes.

The timing of the update is notable. It comes shortly after Google unveiled Gemini Omni, a new multimodal AI model capable of generating highly realistic video content that demonstrates understanding of science, history, culture, and physics. At the same time, generative AI video tools have fueled an increase in fake movie trailers, celebrity deepfakes, and misleading viral clips across social media. In recent months, fake trailers for films like Avengers: Doomsday attracted millions of views online, with many users mistakenly believing they were official releases. Earlier this year, YouTube shut down several major fake trailer channels following investigations into deceptive AI-generated content.

YouTube’s new labels will also become significantly more visible. For long-form videos, labels will now appear directly beneath the video player and above the description section. On YouTube Shorts, labels will appear as overlays directly on the video itself. The company says the goal is to provide “context at a glance” so viewers immediately know when realistic content has been generated or heavily altered by AI. Less realistic or clearly fictional AI content, such as animated fantasy scenes or minor alterations, will continue to have labels hidden within the expanded description section instead.

Importantly, YouTube emphasized that these disclosure labels will not affect monetization or recommendation algorithms. Videos labeled as AI-generated will still be eligible to earn revenue and appear in recommendations like any other content. The labels are intended purely as transparency tools rather than punishments.

Some labels will also become permanent. Videos created using YouTube’s own AI tools, such as Veo or Dream Screen, will automatically retain their AI disclosures. The same applies to videos containing C2PA metadata, an industry standard developed by the Coalition for Content Provenance and Authenticity that verifies when content has been fully generated by AI. Companies including OpenAI, Nvidia, Kakao, and Eleven Labs have recently committed to supporting the C2PA standard.

The update is also part of YouTube’s broader effort to combat deepfakes and unauthorized AI likenesses. Earlier this month, the platform expanded its likeness-detection system to all adult creators, allowing users to scan YouTube for AI-generated face matches and request removal of unauthorized synthetic content. Concerns surrounding deepfakes have intensified in recent years, with public figures like physicist Brian Cox and actor Keanu Reeves speaking out against AI-generated impersonations and misleading content online.

As AI tools continue reshaping online media, YouTube appears determined to balance

innovation with transparency. With millions of videos uploaded every day, automatically identifying synthetic content will be a massive challenge. Still, the platform’s new labeling system signals a growing recognition across the tech industry that viewers increasingly need clear ways to distinguish between authentic footage and AI-generated reality.

Works Cited

Spangler, Todd. “YouTube Will Start Automatically Tagging Videos That Make ‘Significant’ Use of AI, and It’s Making Labels for AI-Generated Content More Prominent.” Variety, 27 May 2026, https://variety.com/2026/digital/news/youtube-ai-video-labels-automatic-detection-1236758865/.

“Youtube Says It’s Making AI Labels More Prominent and Automatic.” Yahoo News Canada, 27 May 2026, https://ca.news.yahoo.com/youtube-says-making-ai-labels-151432090.html.

Yin-Poole, Wesley. “YouTube Announces Automatic AI Video Detection and More Prominent Labels Amid Rise of Fake Movie Trailers and Deepfakes.” IGN, 27 May 2026, https://www.ign.com/articles/youtube-announces-automatic-ai-video-detection-and-more-prominent-labels-amid-rise-of-fake-movie-trailers-and-deepfakes.

“Hero Image.” Mashable, https://helios-i.mashable.com/imagery/articles/077XUN6hs95i9eGndQx4Gs5/hero-image.fill.size_1248x702.v1761145492.jpg. 

“Image.” PCMag, https://i.pcmag.com/imagery/articles/06Vr5fnEtv30iTu05YxKSRl-1..v1699970414.jpg.