“I alone have the right to be me.” YouTube offers Hollywood celebrities and artists a free deepfake detection tool, further developing the fight against identity theft generated by artificial intelligence.
Last month, Google’s video platform launched its image protection tool, which helps identify content in which a face appears altered, or generated, using AI technologies. The project initially targeted government officials, journalists and other political personnel.
The platform is now extending its access to actors and musicians, via talent agencies, management companies and the stars they represent.
The tool allows you to “search for AI-generated content that takes on the appearance of a participant, such as a deepfake of their face, and gives them the power to find it and request its deletion.”
Celebrities and artists can access it even without having a channel on the platform.
“YouTube opening its deepfake detection capabilities to public figures marks a turning point in how platforms approach identity protection in the era of generative AI,” said Alon Yamin, CEO and co-founder of Copyleaks, an AI-generated content detection platform.
“The technology to reproduce a person’s face, voice and facial expressions has advanced faster than the guardrails around it, creating a gap that malicious actors are already exploiting.”
Major issues
The initiative comes at a time when hyper-realistic videos of deceased celebrities are multiplying, created with consumer applications like Sora, the OpenAI tool.
The application triggered a flood of videos of Michael Jackson and Elvis Presley. And last month, OpenAI shut down its Sora app.
Last February, Irish director Ruairí Robinson had already created a strikingly realistic clip showing Brad Pitt fighting with Tom Cruise on a roof, from a two-sentence prompt.
Widely distributed and causing great concern in Hollywood, it was generated with Seedance 2.0, a tool belonging to the Chinese group ByteDance. The Irishman has also created other videos. One shows Brad Pitt confronting a zombie ninja armed with a saber, another features him fighting a robot, flanked by the inevitable Tom Cruise.
Charles Rivkin, head of the Motion Picture Association, the association of major American production companies, called on ByteDance to “immediately cease its counterfeiting activities”, accusing it of violating copyright.
YouTube, for its part, explains that it is working with the main talent agencies to improve the detection of problematic images and better protect artists.
Their heritage
The platform “does the right thing by providing these tools for free to talents, so that they can protect their assets,” rejoices Jason Newman, of the management and production company Untitled Entertainment.
“Their heritage is their face, their body, who they are, what they do, their way of expressing themselves,” he adds in an interview with Hollywood Reporter magazine.
The development of the tool follows complaints from prominent American personalities denouncing the cumbersome procedure on YouTube for reporting and removing a deepfake.
“The stakes are particularly high because deepfakes can be used to spread disinformation, manipulate markets, damage reputations or suggest misleading support. Robust detection is no longer optional,” explains Alon Yamin.
“Detection systems must be extremely accurate, continually updated and combined with clear rules and rapid withdrawal procedures to be effective,” he continues.
“This will not completely eliminate deepfakes, but can significantly reduce their reach and impact, by making it more difficult for manipulated content to circulate without being detected or challenged,” argues the head of Copyleaks
