As advances in artificial intelligence accelerate, talent agencies are beefing up their defenses to protect Hollywood stars from misleading and doctored images and videos that could put them at risk.
The rise of generative AI and “deepfakes” (videos or photos that falsely represent a person’s image) have led to widespread dissemination of unauthorized clips that can harm celebrity brands and businesses.
These clips claim to show celebrities saying or doing things they have never said or done. For example, fake nudes of celebrities or videos created to make it appear that Hollywood stars are endorsing products they don't actually use. And the problem is expected to grow even further.
Now, technological tools that use AI are emerging to combat that threat, and the entertainment industry is under threat.
Talent agency WME has partnered with Loti, a Seattle-based company that specializes in software used to flag fraudulent content posted on the internet, including client likenesses. The company, which has 25 employees, will immediately submit a request to his online platform to remove the infringing photos and videos.
Financial details of the transaction were not disclosed.
In Hollywood, artificial intelligence has been seen as both friend and foe. Artificial intelligence is a tool that has the potential to make processes more efficient and spark new innovations, but it is also seen as a job killer and another way for intellectual property to be stolen.
The need for stronger protections against AI also played a central role in last summer's strikes by the Writers Guild of America and the actors union SAG-AFTRA. The nonprofit Artist Rights Alliance on Tuesday published a petition calling on tech companies to “stop devaluing” their work, with signatures from 200 musicians including Billie Eilish and Elvis Costello. The letter was published. As deepfakes proliferate, government agencies want to use AI to thwart bad actors online.
“The worst kind of whack-a-mole is dealing with the deepfake problem without a technology partner to back you up,” said Chris Jacmin, partner and head of digital strategy at WME.
Loti co-founder Luke Arrigoni launched the startup about a year and a half ago. He previously ran an artificial intelligence company called Arricor AI, and before that he was a data scientist at Creative Artists Agency, WME's main rival.
Arrigoni said Loti started working with WME about four or five months ago. WME's clients provided Lottie with several photos of themselves taken from different angles. Short audio clips are also recorded and used to identify fraudulent content. Loti's software searches her web, reports these fraudulent images to clients, and sends removal requests to the platform.
“There is a growing sense that this is an impossible problem,” Arrigoni said. “There's almost a saying now that people say, 'Once it's on the Internet, it's on the Internet forever.' We're working hard across the company to dispel that myth. Masu.”
Arrigoni declined to discuss the financial terms of the partnership or the number of WME clients using Roti's technology.
Jacmin said that before using Roti's technology, his agency's staff had to combat deepfakes in a more ad hoc manner. We should ask web platforms like YouTube and Facebook to remove abusive material based on what we see while browsing and what we hear from our clients. Client fans flag tampered material.
Loti's technology makes this problem even more visible. Depending on the client's wishes, we may not remove all unauthorized content. But at least the performers will know what's in there.
In 2022, companies like Meta and Google were already dealing with removing billions of ads and ad accounts that violated their deception policies, Jacmin said.
Now, more and more people in Hollywood are concerned about how new AI models trained using some public data may use copyrighted works. These technologies could further blur the line between real and fake.
Harmful and false content left unattended for too long can harm a client's business opportunities and commercial support.
“They are so realistic that it would be difficult for most people to tell the difference,” Arrigoni said.
This is the latest partnership between WME and its parent company Endeavor with an AI company. In January, WME partnered with Chicago-based startup Vermillio to protect clients from IP theft by detecting when generated AI content uses a client's likeness or voice.
Endeavor is a minority investor in Speechify, which develops text-to-speech technology. Endeavor CEO Ari Emanuel used Speechify's tools to create a synthetic version of his own voice as he gave the opening remarks at Endeavor's earnings call last year. (Endeavor announced Tuesday that its largest shareholder, Silver Lake, will take the company private in a deal that values the company at $13 billion.)
So far, Roti has been self-funded, Arrigoni said. He said he invested $1 million into the company himself. The company is currently in the process of raising an undisclosed amount of funding for a seed round.
