YouTube announced on Tuesday that it was being used to identify AI claims (deep forgery) and to expand new “photo-detection” techniques to entertainment practitioners.
The technology operates in a similar way to the existing “Contant ID” content recognition system on YouTube, which detects copyrighted content from user uploads and allows copyright owners to request the deletion or sharing of the benefit of the video. Portrait detection techniques work similarly, but target simulated facials. This function is designed to help protect creators and other public figures from unauthorized use of their identity information. This is quite common for public figures, and surveys indicate that celebrity portraits are frequently used in fraudulent advertising. For the first time in the past year, the technology was made available to some YouTube creators through a pilot scheme, which was further expanded this spring to include politicians, government officials and journalists. YouTube indicated that the technology would now be open to entertainment practitioners, including brokerage companies, management companies and flag artists. YouTube has been supported by key institutions such as CAA, UTA, WME and Unite Management, which provided feedback on the new tool. Using this tool does not require artists to create their own YouTube channel.
The detection function does not directly identify the user’s face, but scans the AI generation content and detects visual matching with the registered user’s face. Users can then choose to request the deletion of video, submit a copyright request for deletion or take no action on the grounds of privacy policy irregularities. YouTube stated that it would not delete all content because, according to its rules, imitation and irony were allowed. YouTube indicated that the technology would also support audio in the future. In this connection, YouTube also advocates similar protection measures at the federal level in support of the Counterfeiting Bill in Washington, D.C. The Act will regulate the use of AI for unauthorized reproduction of individual voices and foreign trade. YouTube has not disclosed how many deletions of deep AI forgery have been processed, but in March it was indicated that the number of deletions was “small”.
