People from the entertainment industry, including talent agencies and management companies as well as the celebrities and creators they represent, can now use YouTube’s likeness detection technology to identify their deepfakes. However, there’s a caveat. Finding deepfakes does not guarantee their removal.
In his Instagram Reel, Rene Ritchie, a Creator Liaison for YouTube, explained how it works:
“The tool proactively looks for participants’ likeness in AI-generated content across the platform. If a match is found, like an AI deepfake of our faces, it will surface that video in Studio, where we can review it and request removal through YouTube’s privacy process. Now, putting my journalist cap back on for just a second, does requesting removal automatically result in a takedown? And the answer is no.”
This is currently an experimental feature and available only to select creators who have to enroll for it by submitting government-issued IDs and a brief selfie video.
“Whether you’re just now getting it for the first time, or you’ve had access through the YouTube Partner Program for a while, but haven’t checked it out yet, just head into YouTube Studio and look for the likeness tab under Content Detection,” Ritchie adds.
The Indian creator contention
India is witnessing a steady rise in deepfakes and voice clones. Last December, a deepfake intimate video of YouTuber and creator Payal Dhare, popularly known as Payal Gaming, had gone viral. Following her complaint, Maharashtra Cyber Police were able to ascertain that the video was a deepfake. This incident sparked a debate about the safety of creators in an increasingly AI-generated world of content.
Notably none of the top Indian creators, including Dhare, have commented on this feature and its importance to our country’s creator economy.
India’s Information Technology (IT) Act is not yet up-to-speed with the rapidly changing AI content realities. With social platforms building identity protection infrastructure, creators are moving closer to owning and protecting their digital presence.
Detection is not enough
YouTube’s likeness detection technology will go a long way in identity protection, but there are a few challenges that creators have to navigate.
The feature can only be used to detect visual matches of an enrolled creator’s face. It does not cover audio detection. An important nuance worth noting is that detection and likeness removal requests do not guarantee removal. Ritchie explains, “YouTube has a long history of protecting free expression, and it specifically looks to protect content that’s in the public interest, like parody and satire, even when it’s used to critique world leaders.”
Working towards a safer digital space
YouTube’s likeness detection technology and its rollout for the creator community is welcome news, but the technology in itself is a work in progress. The platform is still working on ironing out issues that arise as AI gets sharper and more nuanced. Its large-scale adaptation will depend on the accuracy of detection and the subsequent actions taken. For the 900-million strong Indian creator community, it is a promise of a safer online environment and protection of their identity and creative work.
Ritchie says,” YouTube’s absolute priority is on safety and protection, but YouTube is exploring similar future paths for likeness that could open up entirely new revenue opportunities for creators and artists to manage, authorise and benefit from AI likeness.”
Social media platforms, such as YouTube, should not limit their research and development in this direction. They need to expand beyond facial recognition. Voice detection, speedier enforcement protocols and better escalation mechanisms should be implemented on priority.
For creators, their face and voice are assets that can be replicated, misused or monetised without consent. It is critical for them to actively monitor their digital presence and use platform tools to report impersonation early.
