YouTube Launches AI Tool for Celebrities to Identify and Request Removal of Deepfake Videos
YouTube has introduced an AI-powered facial recognition tool allowing celebrities to find and request removal of deepfake videos featuring their likeness, balancing new digital rights protections with content policies.
- • YouTube's AI tool identifies celebrities' faces in videos to monitor and request deepfake removal.
- • Participants must submit ID and a verification selfie to use the system.
- • Video removal is not guaranteed and depends on YouTube's privacy and content policies.
- • YouTube's tool is similar to Content ID but does not yet offer monetization options for deepfake content.
Key details
YouTube has expanded its artificial intelligence capabilities to assist celebrities in monitoring and managing deepfake videos that use their likeness on the platform. This new tool employs facial recognition technology to identify public figures’ faces within videos, enabling them to track and request the removal of unauthorized deepfake content.
Initially tested last fall with content creators, the system was extended earlier this year to include politicians and journalists. Now, celebrities—even those without official YouTube accounts—can participate by submitting identification documents and a selfie verification video.
However, YouTube makes clear that removal requests are subject to its privacy and content policies, particularly protecting videos classified as parody or satire. The platform has acknowledged that creators utilizing this system have generally submitted only a low volume of removal requests.
This facial recognition tool is comparable in purpose to YouTube’s Content ID system, which manages copyrighted music and videos, though it currently does not support monetization or revenue sharing for detected image content. In a related development, YouTube recently introduced the option for creators to digitally clone their appearances for use in videos.
Industry players like the talent agency Creative Artists Agency (CAA) are also adapting by creating biometric databases of their clients’ likenesses, preparing for commercial use in the growing AI-driven content landscape. Some celebrities, including TikTok personality Khabi Lam, have begun exploring this space commercially by selling rights to their digital likenesses. Experts suggest that, with appropriate financial compensation, deepfakes could evolve into a novel way for artists to engage fans.
While the tool offers a new form of protection and control for celebrities over their image rights on YouTube, the removal of deepfake videos remains discretionary and contingent upon complex policy considerations.
This article was translated and synthesized from Brazilian sources, providing English-speaking readers with local perspectives.
Source articles (2)
Source comparison
Testing timeline for AI tool
Sources report different timelines for when the AI tool began testing with content creators.
olhardigital.com.br
"the system was initially tested with content creators last year"
segundabase.com.br
"the feature, which began testing with content creators last fall"
Why this matters: One source states the testing began last year, while the other mentions it started last fall. This discrepancy could affect readers' understanding of how long the tool has been in development.