Conversations about generative artificial intelligence (AI) are making their way through offices, schools, homes, and now courtrooms. It’s so easy: simply prompt an AI bot with a few details to write or draw, and it spits out something resembling an original product. This can help create content more efficiently or even automate some tasks.
But where does the bot get its ideas? AI models need to be trained with original content to be able to produce AI-generated materials. AI runs into ethical and legal obstacles when the original creators of that content don’t give their permission, are not paid fairly, or not paid at all, for this use of their work.
In December 2023, the New York Times sued OpenAI, an artificial intelligence research company that promotes generative pretrained transformers (“GPT”), and Microsoft Corporation. OpenAI had used “millions of articles” from the Times to train its chatbots, with no compensation for the original journalism. The complaint in the lawsuit accuses the tech companies of “avoid[ing] spending the billions of dollars that The Times invested in creating that work by taking it without permission or compensation.”
Comedian Sarah Silverman and several other authors, including Michael Chabon and Ta-Nehisi Coates, also sued OpenAI for direct copyright infringement; vicarious infringement; violation of the Digital Millennium Copyright Act (DMCA) by removing copyright management information; unfair competition; negligence; and unjust enrichment (although several of these claims have been dismissed). The authors contend that OpenAI scraped their works to train the bots and thus illegally copied their work without any compensation.
Creators cannot deny that generative AI is here to stay. AI companies, however, cannot treat creators cavalierly and steal from their work or identities. AI companies must appropriately compensate the creators of the original “human-made” content.
The legal community must also recognize that there is a need for increased governance and regulation in an area that is undergoing extreme growth and expanding applications. Attorneys should reimagine the old tools in spheres such as copyright and right of publicity not only to obtain proper permission and provide appropriate compensation, but also work to create new tools to deal with a new concern.
Recently, the Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA), which works to protect actors' rights, took an important step in dealing with this new reality. They introduced as part of its standard contract, a waiver for performers whose voices are used in dynamic AI-assisted advertising. The agreement has rules regarding consent and compensation, including that the actors must be aware and consent to the use of their voice to create a digital replica, and know which products are going to be advertised, and requires actors be paid for each new use.
AI companies must also be wary of real-world implications, including impersonations of a speaker to gain access to voice-activated accounts, to defraud loved ones, or for phishing.
Pollock Cohen recently filed a class action lawsuit against a tech company called Lovo.ai that alleges that the company’s AI-driven text-to-voice software improperly used, cloned, and sold actors’ voices without their knowledge, permission or compensation. We hope that this case puts existing companies – and startups – on notice that there are ethical and legal consequences to stealing original content from hard-working creators, whose livelihoods are at stake.
For more information on generative artificial intelligence (AI) please contact Anna Menkova at Anna@PollockCohen.com or Steve Cohen at SCohen@PollockCohen.com. For more information on the class action lawsuit against Lovo, please email Lovo@PollockCohen.com.
