A New York federal judge has ruled that a case involving two actors who claim their voices were cloned for generative artificial intelligence (GenAI) narration software can proceed with their state-law claims. However, the judge dismissed the plaintiff’s federal trademark claim and most of their federal copyright claims.

Plaintiffs Paul Lehrman and Linnea Sage claimed in a class action complaint in 2024 that GenAI startup LOVO stole voices to create millions of voice-over productions, without permission or proper compensation.

According to the complaint, the voice-over industry is estimated to generate more than $2 billion annually in the United States, and more than $4 billion annually worldwide.

As the complaint notes,

Voice-overs, commonly known as narrations, are integral to various industries, spanning entertainment, advertising, marketing, education, and other corporate sectors. They serve as indispensable tools across a wide array of mediums, including film and television, commercials, animation, audiobooks, e-learning, corporate presentations, marketing presentations, educational videos, interactive voice response (IVR) systems, podcasting, video games, documentaries, virtual assistants, public announcements, dubbing, radio imaging, audio dramas, museum exhibits, phone systems, language learning, medical narration, corporate training modules, websites, sales pitches, and digital content. They are even used in venture capital funding pitches.

Traditionally, actors are hired to read scripts and are paid a negotiated amount for the use of their voices, often including royalties or residuals.

For example, says the complaint,

Actors like Plaintiffs get paid anywhere from $150 for a short recording for a local television tag, to $2,000 for a one-time sales presentation, to $6,000 minimum for a 13-week run of a TV commercial, or more depending on the project.1 Established actors can typically make much more via residuals, and through daily fees which can exceed $1,000 per day and can involve several weeks of work.

LOVO says it’s trying to “disrupt” the narration business by selling a “text-to-speech subscription service that allows its clients – typically companies – to generate voice-over narrations at a fraction of the cost of the traditional model.”

The plaintiffs claim they were “tricked” into providing voice samples on the Fivrr freelancing platform by people they later learned worked for LOVO.

According to the complaint,

On its website, LOVO represents to potential clients that it has agreements with actors allowing LOVO to utilize those actors’ voices and compensating them appropriately for that use. That may be true with respect to some actors, but it is emphatically not true with respect to Plaintiffs: LOVO had no permission to use Plaintiffs’ voices for training its AI Generator, to promote the LOVO service, or to market voices based on Plaintiffs’ voices. And LOVO never compensated Plaintiffs for any of LOVO’s unauthorized uses of Plaintiffs’ voices.

Lovo called the complaint, “a tale filled with pathos and the woes of artificial intelligence,” and argued that Fiverr’s terms of service granted copyright licenses to platform users like LOVO.

The judge ruled that the plaintiffs made cognizable claims under the New York Civil Rights Law, which protects against the unauthorized use of a person’s voice.

However, the judge dismissed the plaintiff’s federal trademark and false advertising claims under the Lanham Act, saying the plaintiffs’ voices do not primarily serve as source identifiers (as they would for a product) and thus do not serve as trademarks. Also, the judge said the plaintiffs didn’t establish consumer confusion about the source of the voices.

The judge allowed a copyright infringement claim from Sage based on LOVO’s use of her voice in public presentations. However, the judge dismissed claims based on LOVO using the plaintiffs’ voices to train their GenAI model.

The dismissal of the training claims was based on a failure to “explain what training is or how it works, even at a very high level of generality.” The plaintiffs have the opportunity to amend their complaint to remedy that failure.

As we blogged about last year, a proposed bill in Congress aims to regulate the creation and use of digital replicas of human beings, including their image, voice, or visual likeness.

Lawsuits based on “deep fakes” are likely to increase as GenAI becomes even better at creating images of human being that are indistinguishable from reality.

Just like the haiku above, we like to keep our posts short and sweet. Hopefully, you found this bite-sized information helpful. If you would like more information, please do not hesitate to contact us here: https://aeonlaw.com/contact-us/.