Sony Music warns companies to stop training AI on its artists’ content

Sony Music Group, one of the world’s largest record labels, has warned artificial intelligence companies and music streaming platforms not to use the company’s content without explicit permission.

Sony Music, whose artists include Lil Nas against using it for training AI models.

“Unauthorized use” of Sony Music Group content in “the training, development, or commercialization of AI systems” deprives the company and its artists of control and remuneration for those works, according to the letter obtained by Bloomberg News.

Copyright infringement has become a major problem for generative AI, which is used to produce all kinds of content, from text to images and videos. As Hollywood actors and writers went on strike last year to protect their professions from AI, a new generation of startups is producing entire albums of AI-generated music, increasing concerns over livelihoods artists and rekindling tensions with streaming platforms.

Sony Music, along with the rest of the industry, is working to balance the creative potential of this rapidly evolving technology while protecting artists’ rights and its own profits.

“We support artists and songwriters who take the initiative to embrace new technologies to support their art,” Sony Music Group said in a statement Thursday. “However, this innovation must ensure that the rights of songwriters and performers, including copyright, are respected.”

Universal Music Group NV has been particularly outspoken over the past year and a half, going so far as to take the “nuclear option” of removing its entire music catalog from TikTok and suing the startup by IA Anthropic for publishing song lyrics without permission.

After months of impasse, Universal has ended its feud with ByteDance Ltd’s TikTok. earlier this month, with an agreement that ensured both parties’ commitment to “working together to ensure that the development of AI in the music industry will protect human artistic talent and the economy that results from it” . flows to these artists and songwriters.

In another sign of the growing conflict between creators and AI companies, text-to-speech startup Lovo Inc. was hit with a class-action lawsuit on Thursday, alleging the company hijacked actors’ voices and deceptively promoted its product as legally marketing their use. Two voiceover actors seek to represent people whose voices Lovo used without permission or compensation for the “purpose of creating or refining its AI text-to-speech generator” or whose AI-replicated voices were used or sold without appropriate compensation, according to a complaint filed in the U.S. District Court for the Southern District of New York.

In many jurisdictions around the world, particularly in the European Union, copyright owners are encouraged to publicly state that their content should not be used for data mining and data education purposes. AI without specific licensing agreements.

In the United States, the music industry has supported federal legislation in Congress that would protect artists’ voices and images from unauthorized use of AI. In April, Warner Music Group Corp. CEO Robert Kyncl testified before a congressional subcommittee, affirming his support for the “NO FAKES” Act.

“We must ensure a robust free market licensing system to enable the use of copyrighted materials for training AI models and strong legal protection for name, image and copyright rights. voice,” he said in prepared testimony.