YouTube creates service for artists to identify versions of themselves made with AI

YouTube announced, this Tuesday (21), that it will make available a new service within its website so that artists and other public figures – actors, musicians, athletes and content creators – can identify versions of themselves created using artificial intelligence technologies, that is, deepfakes, and request the deletion of this type of video.

A deepfake is a video in which someone’s face is digitally inserted over another person’s, in a pre-recorded recording. Thanks to artificial intelligence, eye and mouth movements are synchronized and often voice generation technologies are also used.

Thus, false content is created, used to spread fake news, manipulate political speeches or even generate pornography. Recently, synthetic versions of Brad Pitt and Tom Cruise starred in a fight scene that never existed in the real world, for example. Videos like this raised Hollywood’s alarm regarding the image rights of its stars.

In an interview with the American magazine The Hollywood Reporter, Mary Ellen Coe, YouTube’s business director, says that it is the platform’s responsibility to combat this type of content.

“We’ve been working on this for a while, since we started thinking about the implications of AI on our platform. Frankly, we haven’t even seen all the possibilities that could arise (from this technology), and we’re working with talent agencies and management companies to ensure that public figures can anticipate and prevent negative things from happening,” says the executive.

Testing began about a year and a half ago, with some of the site’s main YouTubers. This year, the company expanded the service to some politicians, finally making it available to anyone who meets the criteria, whether they have a YouTube channel or not.