YouTube has already launched a new tool to detect deepfakes created with AI within its platform, and the key is that it works in a similar way to Swear ID, but focused on a person’s image and not on the copyright. What’s new now is that this technology is expanding to the world of entertainment to help celebrities, agencies and representatives identify manipulated videos that use their face without permission.
YouTube reinforces its fight against deepfakes
YouTube announced the expansion of its technology similarity detection either “likeness detection”a tool designed to locate AI-generated content that imitates a person’s face in videos published on the platform. The feature will now also be available to celebrities, talent agencies, management companies and entertainment figures, after having gone through a pilot with some creators and previously expanded to journalists, politicians and officials.
The movement does not come by chance, because Deepfakes have already become a serious problem for digital reputationscams and misinformation. Especially for public figures, this type of content can circulate very quickly with false promises, misleading advertisements, or made-up statements that appear real at first glance.
How the new tool works
The mechanics are quite reminiscent of Swear IDthe system with which YouTube detects copyright-protected content within videos uploaded by users. The difference is that here the system does not look for a song, a scene or a recorded clip, but rather visual matches with the face of a person enrolled in the program.
In practice, the tool scans AI-generated content to detect if there is a visible match to the registered participant’s face. If you find a suspicious video, that person or their representative can review it and decide whether to request its removal for a possible violation of privacy policies, file a copyright claim, or simply take no action.
Another important detail is that It is not necessary for the celebrity to have their own YouTube channel to use this protection. This greatly expands the scope of the tool, because it allows actors, musicians or public figures who are not regular creators to also monitor whether their image is being used in deepfakes within the platform.
Besides, YouTube made it clear that it will not automatically remove all detected contentsince its rules continue to allow parody and satire materials. That is, detection is not equivalent to immediate removal, but rather to an alert and review system designed to distinguish between harmful manipulation and uses permitted under its rules.
YouTube wants to put an end to deepfakes
The advertisement shows that YouTube wants to get ahead of a problem that is no longer experimental and now affects creators, celebrities, journalists and public figures in basic. The company itself explained that this technology seeks to prevent a person’s identity from being used without authorization, something especially delicate in fraudulent advertisements and fake videos made with AI.
YouTube confirmed that later the technology will also add support for audio. That means that the platform not only wants to detect simulated faces, but also possible voice clones, one of the most dangerous formats of generative AI when used to deceive users or impersonate identities.
At the political and regulatory level, the company is also supporting in Washington the proposal known as NO FAKES Actan initiative that seeks toRegulate unauthorized use of AI-generated voices and images. In other words, YouTube is not just trying to contain the problem within its platform, but rather push a broader framework to respond to a technology that is advancing faster than many current norms.
Keep reading:
• YouTube without ads: 2 simple methods that work in 2026
• YouTube Top fee raises its prices: how much you will have to pay in each knowing to not see ads
• Fewer ads on YouTube lives, the new decision that favors free users
