Audiobook writers fear Spotify is using their voices to train AI

Starling believes Findaway misused the material the writers and narrators entrusted to him. “This is immoral and illegal,” Starling told WIRED, “Rightholders only have the copyrights for the audio book production, but not claim the voice of the narrator.” She is pausing the release of three upcoming titles she planned to distribute through Findaway.

Interest in automating the art of book telling has increased in recent years for both business and technological reasons. Audio book sales have continued to grow, although book and e-book sales have declined and synthetic speech technology has improved dramatically. A number of tools have emerged that allow anyone to clone voices for synthetic narration with one click, but to develop them further, companies still need vast amounts of data.

In industries like entertainment and gaming, contracts that require voice actors to allow tech companies to train their AI models to generate digital narratives about their work have become more common, says Tim Friedlander, president of US-based National Association of Voice Actors. Adobe, maker of Photoshop and other imaging software, recently began training its own AI algorithms on the work of visual creators, unless they choose not to.

“Voice is what voice actors make a living from,” added Friedlander, “and it literally takes the words out of our mouths without our consent.”

Google started offering free synthetic narration for books in 2020. When Apple announced its own line of digital audiobook narrators in January, the company hoped to eliminate the “cost and complexity” that producing a human-narrated audiobook can entail for small publishers and independent authors. The company’s books app lists titles with AI narration as “narrated by digital voice based on a human narrator.”

Apple has used synthetic speech technology for years, including for the Siri virtual assistant, driving directions, and accessibility features. However, some authors and narrators suspect that audio from their e-books has helped the company improve its technology for the complex task of book narration. The length of audiobooks, the complexity of the material, and the impressive skills of talented speakers make narrating books arguably the greatest challenge for synthetic speech technology.

The application of synthetic voices to books also brings new business and cultural challenges. “Most of the companies developing these AI technologies are in the tech space rather than the entertainment space,” says Love of SAG-AFTRA. “They lack the relationships, history of protection, and reliance on licensing rights that voice actors have come to expect.”

Several authors told WIRED that Findaway has grown into a reliable distributor offering lucrative deals on audiobook listings across multiple platforms. But they also say Findaway often asks people to agree to updated agreements, usually with minor changes, when they log into their accounts. The company added the machine learning clause to its distribution agreements in 2019.

Many suspect they signed the machine learning clause without realizing it. “It’s on me that I didn’t initially notice the addition and its full meaning,” says Laura VanArendonk Baugh, an author living in Indianapolis, Indiana. “But the placement was also a bit sneaky.”

Leave a Reply

Your email address will not be published. Required fields are marked *