The AI-based Likeness Detection system offers celebrities and their teams an easier way to detect deepfake videos and request their removal.
However, this tool is not really new. YouTube initially reserved this technology for a limited number of creators before opening it to profiles at risk of misinformation and impersonation, such as journalists, politicians, and public figures.
Now, Google’s video platform wants to go further by suppressing or blocking any publication that impersonates a celebrity. This approach is similar to Content ID, but applied to the likeness of a famous person.
To do this, celebrities will need to provide an ID document and a reference of their face. YouTube will then automatically analyze new videos uploaded to detect unauthorized use of these images. If a potential match is found, the parties involved can review it and request removal through the privacy procedure.
“With the support of leading talent agencies and management companies, including CAA, UTA, WME, and Untitled Management, we have worked to refine how likeness detection can better serve talents. We are excited that celebrities and artists are now eligible to access this tool, whether they have a YouTube channel or not,” YouTube explained on its blog.
However, Google’s platform clarifies that not all detections will be automatically deleted. YouTube wants to retain oversight to protect satirical or parody videos, as well as general interests (such as journalistic or critical content).
