CNN
—
Actress Judy Garland never recorded her voice reading an audiobook of The Wizard of Oz, but soon you’ll be able to hear her take on the children’s novel that inspired the film.
Earlier this week, AI company ElevenLabs announced that it had digitally produced Include celebrity voiceovers from deceased actors such as Garland, James Dean and Burt Reynolds in the newly launched Reader app. The company said the app takes articles, PDF files, ePub files, newsletters, e-books or any other text on your phone and converts it into voiceovers.
“We deeply respect their legacy and are honored to have their voices on our platform,” said Dustin Blank, Head of Partnerships at ElevenLabs. “Adding them to our growing roster of speakers is an important step forward in our mission to make content accessible in every language and voice.”
The company said it has signed contracts with the estates of the actors whose voices are used, but did not disclose details about the compensation. The project shows the potential of artificial intelligence for Hollywood, but also sets a precedent for licensing and working with estates. It also comes at a time when technology has come on leaps and bounds, particularly in the ability to create images, text, and sound, making it easy for anyone to create a version of a person’s voice saying something that person never said.
This, in turn, raises questions in creative industries such as journalism and film about how artificial intelligence can – or even should – be used.
ElevenLabs made headlines earlier this year when its tool was reportedly used to make a fake automated call from President Joe Biden urging people not to vote in the New Hampshire presidential primary.
Copyright issues and authenticity
The partnership with the stars’ estates comes two months after ChatGPT maker OpenAI came under fire after introducing a synthetic voice that bore an eerie resemblance to Scarlett Johansson’s character in the film “Her.” Johansson said in a statement shared with CNN that she was “shocked, angry and in disbelief” that the company would use her similar image. after turning down a partnership opportunity with OpenAI.
Although a person cannot copyright their own voice, it is possible to copyright a recording, according to David Gunkel, a professor in the Department of Communication at Northern Illinois University who studies AI in media and entertainment. The AI is trained on old recordings, and those recordings are subject to copyright.
“ElevenLabs’ “New partnerships are all within the bounds of what the law allows,” he said. “An estate will receive a significant amount of money through licensing and agreements. It’s not unlike a company negotiating a copyright deal to use a popular Queen song in an advertisement. The record company could theoretically say no, no matter how much money it’s offered.”
Bern Elliot, vice president and analyst at research firm Gartner, said AI models can now be trained on fewer audio recordings. Very little is needed to capture tone of voice, speech patterns and other elements – whether it’s a celebrity or a regular person.
“The bigger concern is determining what the owner of these recordings can and cannot do to make money from the voice,” he said.
Media companies are also increasingly using AI for voiceovers. Last week, NBC announced it will bring an AI version of famed sportscaster Al Michaels back to the Olympics this summer, in daily recaps on its streaming platform Peacock. An NBC spokesperson told CNN that Michaels will be compensated for his involvement.
However, it is unclear how AI versions of well-known voices will be received by mass audiences and whether there will be doubts about their authenticity.
“We don’t yet know the market for such things, but you can already see with audiobooks that those read by well-known voices and celebrities are in high demand,” said Gunkel. “If there is a way for celebrities to create all kinds of content without doing the voice-overs themselves, that could open up the market even further.”
