YouTube is broadening its “likeness detection” technology, designed to identify AI-generated content, including deepfakes, specifically for professionals in the entertainment sector. This initiative was announced on Tuesday, with the aim to safeguard creators and public figures from unauthorized use of their likenesses, a recurring issue for numerous celebrities targeted in fraudulent advertisements.
The likeness detection system operates similarly to YouTube’s established Content ID system, which flags copyright-protected material uploaded by users, enabling rights holders to either request removal or share in the revenue generated by those videos. In contrast, the likeness detection feature focuses on identifying simulated faces, thus providing an essential protective mechanism for those whose images may be misappropriated.
After a pilot program last year that included a limited selection of YouTube creators, the tool has since expanded to encompass politicians, government officials, journalists, and now members of the entertainment industry such as talent agencies and management firms, including notable agencies like CAA, UTA, WME, and Untitled Management, which have contributed valuable feedback regarding the new technology.
Importantly, entertainers don’t need to operate their own YouTube channels to use the likeness detection tool. The system scans for AI-generated content that visually matches the faces of enrolled participants, allowing users to request content removals for privacy violations, submit copyright claims, or choose to take no action. YouTube clarifies that it won’t remove all flagged content, allowing parody and satire to remain under its guidelines.
Looking ahead, YouTube intends to expand this technology to include audio recognition as well. Alongside this, the company has been advocating for legal protections at the federal level, supporting the NO FAKES Act, which aims to regulate the use of AI in creating unauthorized reproductions of individuals’ voices and likenesses.
As of now, YouTube has not disclosed the number of removals executed through the likeness detection tool, though it indicated in March that the removals counted so far had been “very small.” This further highlights the ongoing challenges of dealing with AI-generated imitations and the potential need for enhanced regulatory measures as technology continues to evolve.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
